JP7601565B2 - Diagnostic device and diagnostic method for rotating machine system - Google Patents

Description

The present invention relates to a condition diagnosis technology for rotating machine systems.

A rotating machine system drives a rotating machine by transmitting driving force from an electric motor (in other words, a motor) through a power transmission mechanism including a belt, based on AC power, for example. For example, there are many belt-driven rotating machines, such as air compressors, injection molding machines, and fans.

When deterioration or an abnormality occurs in the power transmission mechanism of a rotating machine system, the transmission of driving force becomes insufficient or abnormal. For this reason, technology that can diagnose and detect the condition of the power transmission mechanism is being investigated.

An example of prior art related to the diagnosis of the condition of a rotating machine system is International Publication No. 2018/109993 (Patent Document 1). In Patent Document 1, as an "abnormality diagnosis device for a power transmission mechanism", it is described that the abnormality diagnosis device includes an abnormality diagnosis unit that determines an abnormality in the power transmission mechanism and a current detector connected to the power source of the electric motor, and the abnormality diagnosis unit includes an analysis unit that analyzes the current transmitted from the current detector, and an abnormality determination unit that determines an abnormality in the power transmission mechanism from the analysis result of the analysis unit.

International Publication No. 2018/109993

The condition of the power transmission mechanism, including belts, in a rotating machine system must be properly managed through regular maintenance to ensure the performance and safety of the rotating machine. However, compared to other equipment, equipment that operates continuously for long periods of time or has complex mechanisms incurs relatively higher downtime costs due to regular maintenance.

In an example of Patent Document 1 relating to the diagnosis of the condition of a power transmission mechanism, it is described that the current value acquired from a current detector is frequency-analyzed, and the number of sideband wave peaks other than the rotation frequency of the motor that exceed a threshold is detected, thereby judging whether an abnormality has occurred in the power transmission mechanism.

Conventionally, there has been technology for diagnosing the condition of a power transmission mechanism in a rotating machine assuming a certain load, such as the example in Patent Document 1. In such technology, the condition of the power transmission mechanism is determined using a set, fixed threshold value.

However, in a rotating machine connected to an electric motor and a power transmission mechanism, when the load for driving the rotating machine fluctuates, a method of diagnosis using a fixed threshold value makes it difficult to determine the state, and there is a possibility of misdiagnosis or false alarms. More specifically, when the load on the electric motor changes, the frequency spectrum peak value in the electric information of the electric motor also changes. As a result, a method of determining the state by comparing the electric information with a fixed threshold value may result in misdiagnosis or false alarms due to erroneous determination. An erroneous determination may be, for example, determining that something is normal but is abnormal, or determining that something is normal but is abnormal. Therefore, conventional state diagnosis techniques are difficult to apply to rotating machine systems where the load changes from moment to moment.

The object of the present invention is to provide a technology for diagnosing the condition of a rotating machine system, which can diagnose abnormal conditions, etc., related to the power transmission mechanism with high accuracy while suppressing or preventing erroneous diagnosis, etc., even when there are fluctuations in the load on the electric motor and the rotating machine.

A representative embodiment of the present invention has the following configuration. A diagnostic device for a rotating machine system in one embodiment includes a motor driven based on power supplied from a power source, a power transmission mechanism connected to the motor, a rotating machine connected to the power transmission mechanism, and a diagnostic device connected to the motor or a power line from the power source to the motor and configured to diagnose the state of the power transmission mechanism. The power transmission mechanism includes a first pulley that outputs the rotational driving force of the motor, a second pulley that drives the rotating machine, and a drive transmission component that connects the first pulley and the second pulley. When the load applied to the rotating machine from the motor through the power transmission mechanism fluctuates, the diagnostic device detects the current in the power, normalizes the current value on the time series of the detected current at predetermined time intervals, determines the state of the power transmission mechanism using the normalized current data, and stores and outputs the diagnosis result.

A representative embodiment of the present invention provides a technology for diagnosing the condition of a rotating machine system, which can accurately diagnose abnormal conditions, such as those related to the power transmission mechanism, even when there are fluctuations in the load on the electric motor and rotating machine, while suppressing or preventing erroneous diagnosis, etc.

1 is a diagram showing a configuration of a rotating machine system including a diagnosis device according to a first embodiment of the present invention; FIG. 2 is a diagram illustrating a configuration example of a diagnostic device in the first embodiment. FIG. 2 is a diagram showing a flow of main processing of the diagnostic device in the first embodiment. 5A to 5C are diagrams showing detected current data and current information in the first embodiment; 11 is a diagram showing envelope detection and current normalization in the first embodiment. FIG. 5A to 5C are diagrams illustrating an example of a frequency analysis result of a current in the first embodiment. FIG. 4 is a diagram showing an example of a display screen in the first embodiment. FIG. 4 is an explanatory diagram showing an example of load fluctuation in the first embodiment. FIG. 13 is a diagram illustrating an example of determination in the case of a load fluctuation, in a comparative example of the first embodiment; FIG. 11 is a diagram illustrating an example of determination in the case of a load fluctuation in the first embodiment. FIG. 11 is a diagram showing a flow of main processing of a diagnostic device according to a second embodiment of the present invention. FIG. 11 is a diagram showing an example of a display screen in the second embodiment. FIG. 13 is a diagram showing the Mahalanobis distance in the third embodiment of the present invention. FIG. 11 is a diagram showing a configuration of a rotating machine system including a diagnosis device according to a fourth embodiment of the present invention. FIG. 13 is a diagram showing a configuration of a rotating machine system including a diagnosis device according to a fifth embodiment of the present invention. FIG. 13 is a diagram showing a configuration of a rotating machine system including a diagnosis device according to a sixth embodiment of the present invention. FIG. 23 is a diagram showing an example of a spectral peak in the sixth embodiment.

The following describes in detail an embodiment of the present invention with reference to the drawings. In principle, the same parts are given the same reference numerals in all the drawings, and repeated explanations are omitted.

<First embodiment>
1 to 7, a diagnostic device and a diagnostic method for a rotating machine system according to a first embodiment of the present invention will be described. The diagnostic method according to the first embodiment is a method having steps executed by the diagnostic device according to the first embodiment.

[Rotary Mechanical Drive System]
FIG. 1 shows a configuration of a rotating machine system 100 as a system including a diagnostic device 10 of the first embodiment. The rotating machine system 100 includes the diagnostic device 10 of the first embodiment, an AC power source 1, a switch 2, an electric motor 3, a power transmission mechanism 6, and a rotating machine 4. A portion 101 is a portion configured by connections between the electric motor 3, the power transmission mechanism 6, and the rotating machine 4. The power transmission mechanism 6 is a belt drive transmission unit, and has, as components, a first pulley 3P which is a motor-side pulley, a belt 5, and a second pulley 4P which is a rotating machine-side pulley. The rotating machine system 100 is mainly configured by the AC power source 1, the switch 2, the electric motor 3, the power transmission mechanism 6, and the rotating machine 4, and the diagnostic device 10 is connected to the main part. The diagnostic device 10 may be disposed close to the electric motor 3, etc., or may be disposed remotely. A user U1 who is an operator operates and manages the diagnostic device 10, and checks the state of the rotating machine system 100, etc.

In the first embodiment, an example is shown in which the part 101 including the rotating machine 4 is applied to an air compressor, but the present invention is not limited to this and can be similarly applied to other types of rotating machine systems.

The AC power source 1 is a three-phase AC power source, and supplies power to the electric motor 3 via the switch 2. The switch 2 is a switch board that can switch between a power-on state (in other words, an on state) and a power-off state (in other words, an off state), and may include not only a switch circuit but also other control circuits. The switch 2 is provided midway through the power line 1L that is connected between the AC power source 1 and the electric motor 3. The power line 1L is a line that transmits three-phase AC power. When the rotating machine system 100 starts operating, the user U1 operates the switch 2 to turn on the power. As a result, power from the AC power source 1 is supplied to the electric motor 3 via the power line 1L and the switch 2, and the electric motor 3 is rotated.

The electric motor 3 generates a rotational driving force for the rotating machine 4 based on electric power as motive power. For example, an induction motor is applied to the electric motor 3. However, this is not limiting, and various electric motors can be applied to the electric motor 3. The electric motor 3 is connected to the rotating machine 4 via a power transmission mechanism 6. The rotating machine 4 receives the rotational driving force of the electric motor 3 from the power transmission mechanism 6 and is driven to rotate.

The electric motor 3 has a first pulley 3P provided on the drive shaft (also called the motor shaft). The rotating machine 4 has a second pulley 4P provided on the drive shaft. Each pulley (3P, 4P) is a component that transmits power between the drive shaft and the belt 5. The belt 5 is an example of a power transmission component. The power transmission component is not limited to the belt 5, and a chain or a rope may be applied. The belt 5 is hung between the first pulley 3P and the second pulley 4P with friction or the like. One end of the belt 5 is hung on the first pulley 3P, and the other end is hung on the second pulley 4P. This constitutes a belt drive transmission section (in other words, a winding transmission device) by the pulleys (3P, 4P) and the belt 5.

The diagnostic device 10 is a device that has the function of diagnosing the state of the power transmission mechanism 6, and can be implemented, for example, on a PC or an electronic circuit board. In the configuration example of FIG. 1, the diagnostic device 10 is connected to the power line 1L between the AC power source 1 and the electric motor 3. In particular, the diagnostic device 10 is connected to one point (one of the three-phase lines) on the power line 1L between the AC power source 1 and the switch 2 via a terminal 1N and a detection wire 1S.

The diagnostic device 10 has an electrical information detection unit 11, a diagnosis unit 13, a display device 14, an alarm device 15, an operation unit 16, and a setting unit 17. A user U1 operates the operation unit 16 to input instructions and information to the diagnostic device 10. The operation unit 16 may be, for example, a keyboard, a mouse, an operation panel, or a PC connected for communication.

The electric information detection unit 11 is connected to one point (one of the three-phase lines) of the power line 1L between the AC power source 1 and the motor 3 through the terminal 1N and the detection wire 1S. The electric information detection unit 11 is a part that detects, senses, or measures the electric information of the power line 1L between the AC power source 1 and the motor 3, and is composed of a device such as a circuit. In the first embodiment, a current (phase current in this example) is used as the electric information. That is, in the first embodiment, the electric information detection unit 6 has a current sensor and can detect a current (in other words, current information, current value). The electric information detection unit 11 outputs a signal of the detected current as a current signal E1 to the diagnosis unit 13. The current signal E1 is an analog signal that represents a current value in a time series. Note that an analog-to-digital converter (ADC) is included in the diagnosis device 10, for example, in the diagnosis unit 13 (or in the electric information detection unit 11). The analog signal that is the current signal E1 is appropriately converted into a digital signal by the ADC.

The diagnostic unit 13 is a part that diagnoses the state of the power transmission mechanism 6, particularly the state related to the drive transmission by the belt 5, based on the current signal E1, and performs processes such as storing and outputting the diagnosis results. In other words, the diagnostic unit 13 is a part that detects abnormalities, deterioration, or signs of problems related to the transmission or drive force of the belt 5, and outputs a notification to the user U1, etc. A sign refers to a sign of an abnormality or failure before it actually occurs.

As a modified example, the diagnostic unit 13 may perform some kind of drive control of the part 101 including the electric motor 3 and the rotating machine 4 as output control based on the diagnosis result. For example, this control may include temporarily suspending the drive operation of the electric motor 3 and the rotating machine 4 or decelerating the operation when the diagnosis result indicates an abnormal state or a state of high degree of degradation. In this case, a control signal is given from the diagnostic device 10 to the switch 2 or the electric motor 3, etc.

In the part 101 consisting of the electric motor 3, the power transmission mechanism 6, and the rotating machine 4, the current frequency (sometimes called the electrical frequency, the rotational frequency, etc.) of the electric motor 3 is fm [Hz], and the rotational frequency of the rotating machine 4 is fc [Hz]. The current frequency fm is a frequency corresponding to the rotational speed of the electric motor 3.

The current frequency fm can be known in advance from the frequency of the AC power supplied from the AC power source 1 to the electric motor 3. However, it is not limited to this, and a specific sensor may be provided in the electric motor 3 or between the AC power source 1 and the electric motor 3, and the current frequency fm may be detected by the sensor, or the current frequency fm may be calculated from the sensor detection information.

In the first embodiment and the like, the rotating machine 4 is a rotating machine in which the load may fluctuate, and periodic load torque and rotational fluctuations may occur. Correspondingly, the load of the electric motor 3 may also fluctuate. In this rotating machine 4, the current waveform that changes with the current frequency fm of the electric motor 3 is observed modulated at the rotational frequency fc by the periodic load torque of the rotating machine 4. In other words, the amplitude of the current waveform at the current frequency fm of the electric motor 3 is modulated by the rotational frequency fc of the rotating machine 4.

In the first embodiment, the state of the power transmission mechanism 6, particularly the state related to the transmission of the belt 5, is diagnosed by distinguishing at least two states, "normal" and "abnormal". In terms of information processing, for example, the "normal" state is defined as value 0, and the "abnormal" state is defined as value 1. The "normal" state related to the transmission of the belt 5 corresponds to a case where no slippage or the like occurs between the first pulley 3P, the belt 5, and the second pulley 4P, or where no damage or the like occurs to the belt 5 or the like. The "abnormal" state corresponds to a case where slippage or the like occurs between the first pulley 3P, the belt 5, and the second pulley 4P, or where damage or the like occurs to the belt 5 or the like.

Without being limited to this, the state of the power transmission mechanism 6 may be defined in more detail by three or more state values. For example, a number of state values representing the degree of deterioration between normal and abnormal may be used as state values for the belt 5. For example, a state in which the degree of deterioration is small may be value 1, a state in which the degree of deterioration is medium may be value 2, and a state in which the degree of deterioration is large may be value 3. In this case, the diagnosis unit 13 uses a method for determining the degree of deterioration.

In addition, as a state value, a "premonitory" state may be used as a state prior to reaching the "abnormal" state, in addition to the "abnormal" state. In this case, the diagnosis unit 13 uses a method for determining the premonitory state. There are no limitations on the details of the method for determining the degree of deterioration and the premonitory state.

In the first embodiment, a current sensor is used to diagnose the condition, but this is not limiting, and other types of sensors (such as a microphone) installed near the power transmission mechanism 6 may also be used in combination.

Assuming that the state of the belt 5 of the power transmission mechanism 6 is normal, the relationship between the current frequency fm and the rotation frequency fc can be expressed by the following formula 1. R1 is the diameter of the first pulley 3P. R2 is the diameter of the second pulley 4P. Rm is the number of pole pairs determined by the structure of the electric motor 3, and is a coefficient that converts between the rotation frequency of the mechanical motor shaft and the electrical frequency of the electric motor 3.
Formula 1: fc=R2/R1×fm/Rm

The diagnosis unit 13 first calculates the first information S1, as shown in the speech bubble in FIG. 1, based on the current signal E1 as electrical information acquired from the electrical information detection unit 11. The first information S1 is a status value that indicates the degree of abnormality in the transmission of the belt 5. The diagnosis unit 13 stores and outputs data including the first information S1 as a result of the diagnosis and detection. The diagnosis unit 13 displays, for example, the data including the first information S1 on the display screen of the display device 14. In this way, the status of the belt 5 of the power transmission mechanism 6 is conveyed to the user U1.

The diagnostic unit 3 also determines second information S2 related to the diagnosis result and notifications, etc., according to the first information S1, and controls the output of the notification device 15, etc., according to the value of the second information S2. That is, the diagnostic unit 13 provides a notification signal to the notification device 15 according to the value of the second information S2, thereby notifying the user U1 of a warning, etc. The notification device 15 issues a warning to inform the user U1 of an abnormal state of the belt 5, etc., according to the notification signal.

The diagnostic unit 13 determines the state value of the first information S1, for example, based on a comparison between a spectrum peak value described below and a set threshold value (defined as H). For example, if the spectrum peak value is equal to or greater than the threshold value H, it is determined to be in an abnormal state (in other words, a state in which the degree of abnormality is sufficiently high). The diagnostic unit 13 also determines the second information S2 according to the state value of the first information S1.

The display device 14 is built into the diagnostic device 10 or is externally attached, and displays images on the display screen. The diagnostic device 10 displays information on the display screen of the display device 14 using a specified graphical user interface. The notification device 15 is a device that realizes notification to the user U1 using sound, light emission, screen display, etc. The notification device 15 may be, for example, a buzzer device that generates a buzzer sound or the like and outputs it from a speaker, or a device that outputs an abnormal condition from a speaker using synthesized voice. The notification device 15 may be a device that conveys the condition by emitting a lamp, a device that displays a warning or the like on a display screen, or a device that notifies a destination of the condition by email or the like. The notification device 15 and the display device 14 may be integrated.

The diagnostic device 10 diagnoses the condition of the power transmission mechanism 6 while the part 101 including the electric motor 3, power transmission mechanism 6, and rotating machine 4 is in operation, i.e., from the start to the end of operation. For example, the diagnostic device 10 can determine the operating state of the part 101 from the detection state of the electrical information detection unit 11, and may control the start and end of the diagnosis in accordance with that operating state. Alternatively, the diagnostic device 10 may input information about the operating state from the electric motor 3, etc., and control the start and end of the diagnosis in accordance with that information. Alternatively, the diagnostic device 10 may control the start and end of the diagnosis based on the operation of the operation unit 16 by the user U1.

The setting unit 17 can set setting information such as the threshold value H related to the processing of the diagnosis unit 13 based on the user U1 checking the screen of the display device 14 and operating the operation unit 16.

In the configuration example of FIG. 1 of the first embodiment, the diagnostic device 10 is installed externally to the part 101 including the rotating machine 4, but as described below, this is not limited to this. The diagnostic device 10 may also be implemented as built into the electric motor 3, the switch 2, etc.

[Rotating Machinery]
In other words, the portion 101 including the rotary machine 4 is an industrial machine that constitutes a manufacturing line or the like. A configuration example of a compressor, particularly an air compressor, as a specific example of the rotary machine 4 is as follows. There are various types of air compressors, and any of them can be applied. In the case of an air compressor, in the portion 101 of FIG. 1, the rotary machine 4 is composed of an air compression mechanism. Power generated by the electric motor 3 is transmitted to the air compression mechanism by the power transmission mechanism 6. The air compression mechanism includes an intake mechanism and a discharge mechanism. A rotor or a piston is provided inside a container of the air compression mechanism. Air sucked in from the intake mechanism is supplied to the container of the air compression mechanism, and the air is compressed by the air compression mechanism. The air compression mechanism compresses the air in the container by the rotational motion of the rotor or the reciprocating motion of the piston in the container based on the power transmitted from the power transmission mechanism 6. The air compressed by the air compression mechanism is discharged from the discharge mechanism. The compressed air discharged from the discharge mechanism is supplied through piping to a device that requires the compressed air (for example, an air cylinder or an air motor).

For example, the load on the air compressor may vary depending on the pressure of compressed air required for the device. The current supplied to the electric motor 3 and the rotational driving force supplied from the electric motor 3 to the air compression mechanism, which is the rotating machine 4, through the power transmission mechanism 6 vary depending on the load fluctuation.

The applicable rotating machine 4 is not limited to an air compressor. The rotating machine 4 may be any machine that receives a rotational driving force from a power transmission mechanism 6. When the rotating machine 4 is, for example, a press machine, an example of the configuration is as follows. A press machine is a machine that processes materials such as metals by applying strong pressure and deforms them into a shape based on a die. In the case of a press machine, in part 101 of FIG. 1, the rotating machine 4 is composed of a press mechanism. The power generated by the electric motor 3 is transmitted to the press mechanism by the power transmission mechanism 6. The press machine converts the rotational motion of the electric motor 3 into a reciprocating motion in the press mechanism through the power transmission mechanism 6. The press mechanism performs press processing on the supply material in the die by the reciprocating motion. For example, the load required by the press machine may vary depending on the processing content in the press mechanism. Depending on the load fluctuation, the current given to the electric motor 3 and the rotational driving force given from the electric motor 3 to the press mechanism, which is the rotating machine 4, through the power transmission mechanism 6, vary.

[Diagnostic Device]
FIG. 2 shows an example of the hardware and software configuration of the diagnostic device 10. The diagnostic device 10 includes a processor 201, a memory 202, an auxiliary storage device 203, a communication interface device 204, a current sensor 205, a display device 14, and an alarm device 15, which are connected to each other via a bus or the like. The processor 201 is composed of a CPU, a ROM, a RAM, and the like, and is a controller, and realizes a state diagnosis function based on the functional block configuration of FIG. 1 based on software program processing. A part of the processing may be realized by a dedicated circuit. The memory 202 or the auxiliary storage device 203 stores a control program 210, setting information 220, processing data 220, and the like. The auxiliary storage device 203 stores diagnostic result data 240, and the like. The processing data 220 and the diagnostic result data 240 include the first information S1 and the second information S2 of FIG. 1. The processing data 220 includes a spectrum analysis result, an envelope signal, a normalized signal, and the like, which are used or generated during processing, as described below. The communication interface device 204 performs communication processing with the outside through a predetermined communication interface. The current sensor 205 constitutes the electric information detection unit 11 in FIG.

[Processing flow (1)]
3 shows a flow of main processes performed by the diagnostic device 10, particularly the diagnostic unit 13. The flow in FIG. 3 has steps S101 to S114. In step S101, the diagnostic unit 13 sets a flag F for the diagnostic operation. The value 0 of the flag F indicates stop, and the value 1 indicates start. For example, when the user wishes to start a diagnostic operation, the flag F may be set to a value of 1 based on a button operation on the operation unit 16. Alternatively, when the diagnostic device 10 has a timer function, the flag F may be automatically set to a value of 1 at preset intervals using the timer function.

In step S102, if the value of flag F is 1 (i.e., start), the diagnostic unit 13 proceeds to step S103, and if the value is 0 (i.e., stop), the diagnostic unit 13 ends the flow. In step S103, the diagnostic unit 13 converts the analog signal, which is the current signal E1 from the electrical information detection unit 11, into a digital signal using an ADC. In other words, this digital signal is the detected current data in a time series.

In step S104, the diagnostic unit 13 performs an operation of acquiring and collecting unit time data DU as data for each predetermined unit time TU (in other words, sampling time) based on the above-mentioned detected current data. In other words, the unit time data DU is a sampling current value. The diagnostic unit 13 repeats this operation multiple times for each predetermined accumulation time (TL).

Figure 4 shows the relationship between the detected current data and various times. In the graph of Figure 4 (A), the horizontal axis is time (time point t) and the vertical axis is current [A (amperes)]. Time 401 indicates a time consisting of accumulation times TL1, TL2, ..., TLm as multiple (m) accumulation times TL. Note that statistical values such as the average described below may be calculated in units of one accumulation time TL, or in units of a predetermined multiple (m) accumulation times TL as one time 401.

The graph in FIG. 4B shows an enlargement of one accumulation time TL. One accumulation time TL includes multiple unit times TU. Current information Ir (in other words, current value) is included as unit time data DU for each unit time TU. That is, one accumulation time TL includes multiple pieces of current information Ir, which are multiple unit time data DU. For each piece of data for each accumulation time TL, the envelope signal, normalized signal, spectral peak, etc., described below, are calculated.

In step S105, the diagnostic unit 13 accumulates each unit time data DU (current information Ir) during the accumulation time TL collected in step S104 in a memory (memory 202 in FIG. 2). In the time series, the process of collecting and accumulating data is repeated in the same manner for each accumulation time TL.

[Process flow (2)—Envelope detection process]
In step S106, the diagnosis unit 13 performs an envelope detection process on the current information Ir for the accumulation time TL. Fig. 5 shows the envelope detection process in step S106. In step S106, the diagnosis unit 13 inputs the current information Ir, which is the unit time data DU accumulated in step S105, in a time series. The diagnosis unit 13 performs an envelope detection process based on the current information Ir and extracts it as an envelope signal IhL.

(A) in FIG. 5 shows an example of the envelope signal IhL. In the graph in (A), the horizontal axis is time and the vertical axis is current. Line 501 is an example of the envelope signal IhL. In (A), an envelope signal for data corresponding to one accumulation time TL in FIG. 4 is illustrated. Such an envelope signal IhL is a wave that changes periodically as shown, and rotation information on the rotating machine 4 side, particularly the rotation frequency fc, is expressed. By such envelope detection, the diagnosis unit 13 can extract rotation information on the rotating machine 4 side (i.e., information corresponding to the rotation frequency fc) from the detected current of the electric motor 3. The envelope signal IhL within the accumulation time TL has a maximum value and a minimum value. The maximum value is max(IhL) and the minimum value is min(IhL). As an example, a maximum value 502a and a minimum value 502b are shown.

In the example of the first embodiment, the diagnosis unit 13 performs heterodyne detection in the envelope detection of step S106. In heterodyne detection, the current information Ir is multiplied by a sine wave function that changes with the current frequency fm of the motor 3, as shown in the following equation 2. The diagnosis unit 13 generates the sine wave function using the current frequency fm of the motor 3. A multiplier in the diagnosis unit 13 multiplies the current information Ir by the sine wave function. The current information after the heterodyne detection of the current information Ir is set as post-detection current information Ih.
Formula 2: Ih=Ir×sin(2π×fm×t)

Here, the current information Ir will be described assuming that it follows the following formula 3. In formula 3, Io is the current amplitude. From formula 3, formula 2 can be transformed into formula 4 using the half-angle formula.
Formula 3: Ir={Io×sin(2π×fc×t)}×sin(2π×fm×t)
Equation 4: Ih={Io×sin(2π×fc×t)}×(1−cos(2×2π×fm×t)/2

The diagnosis unit 13 passes the post-detection current information Ih of Equation 4 through a low-pass filter (LPF) that removes the frequency component of 2×fm [Hz]. Then, an envelope signal IhL that changes with the rotational frequency fc of the rotating machine 4 is obtained by the following Equation 5.
Equation 5: IhL={Io×sin(2π×fc×t)}/2

As described above, the diagnosis unit 13 converts the current information Ir in FIG. 4 into an envelope signal IhL expressed by the rotation frequency fc as shown in FIG. 5(A) by the envelope detection process in step S106. In this example, heterodyne detection is applied during envelope detection, but this is not limiting, and Hilbert transform or the like can also be applied. Each calculation of the diagnosis unit 13 may be realized by software program processing or by a hardware circuit.

[Process flow (3)-normalization process]
3, the diagnostic unit 13 performs normalization processing of the electrical information (particularly, current information). In step S107, the diagnostic unit 13 performs normalization processing (particularly, envelope signal normalization processing) on the envelope signal IhL obtained in step S106 to obtain an envelope normalized signal Ist. The "normalization" used in the first embodiment is mathematical normalization, and is conversion to a value within a predetermined range (for example, a range from 0 to 1).

5B shows an example of the envelope normalized signal Ist after the normalization process in step S107. A line 503 is an example of the envelope normalized signal Ist. The diagnosis unit 13 uses the following equation 6 to calculate the envelope normalized signal Ist. The envelope normalized signal Ist is a function of time (t) and is represented by Ist(t). The calculation of equation 6 is a normalization calculation in which the difference between the current value and the minimum value of the envelope signal IhL is divided by the difference between the maximum value and the minimum value of the envelope signal IhL.
Equation 6: Ist(t)={IhL(t)-min(IhL)}/{max(IhL)-min(IhL)}

By calculating Equation 6, the envelope signal IhL (line 501) in FIG. 5A is normalized to the envelope normalized signal Ist (line 503) in (B). In (B), range 504 is the range from the minimum value 0 to the maximum value 1 in normalization. In other words, the envelope normalized signal Ist is data normalized to the range from the minimum value 0 to the maximum value 1.

In this way, the diagnostic unit 13 repeats the process including normalization for each detected current data within each accumulation time TL. The envelope normalized signal Ist (Ist1, Ist2, ...) is obtained for each accumulation time TL (TL1, TL2, ...) in Figure 4.

[Process flow (4)—Spectral peak detection process]
3, the diagnostic unit 13 performs a spectrum analysis process using a fast Fourier transform (FFT) on the envelope normalized signal Ist for each accumulation time TL to obtain a spectrum that is a result of the spectrum analysis. In the next step S109, the diagnostic unit 13 performs a spectrum peak detection process based on the spectrum obtained in step S108. The diagnostic unit 13 detects a spectrum peak value (denoted as P) for each accumulation time TL.

Figure 6 shows the spectrum peak detection process of step S109. The graph in Figure 6 shows a spectrum, with the horizontal axis being frequency [Hz] and the vertical axis being amplitude [dimensionless quantity]. Figure 6 (A) shows an example in which the state of the belt 5 of the power transmission mechanism 6 is normal, and Figure 6 (B) shows an example in which the state is abnormal. In (A), the spectrum peak 601 shows a spectrum peak in a normal state, and is represented by a certain frequency and amplitude value. The frequency of the spectrum peak 601 shows the rotational frequency fc of the rotating machine 4 as the rated frequency. The amplitude value of the spectrum peak 601 is shown by the amplitude value a1. In the envelope signal IhL of the above-mentioned step S106, the rotational frequency fc, which is the rotation information on the rotating machine 4 side, is extracted. As a result, the frequency of the spectrum peak 601 in the normal state of (A) coincides with the frequency band of the rotational frequency fc.

On the other hand, in (B), the spectrum peak 602 in an abnormal state is represented by a certain frequency and amplitude value. In this example, the frequency of the spectrum peak 602 in an abnormal state is at the same position as the frequency of the spectrum peak 601 in a normal state, i.e., the rotation frequency fc. In the spectrum in an abnormal state, as shown in the figure, a sideband 603 occurs at a different frequency position near the rotation frequency fc of the spectrum peak 602. The amplitude value of the spectrum peak 602 is indicated by the amplitude value a2.

The detection range (corresponding frequency range) of the spectrum peak in step S109 may be a preset range, which may be the range of the rotational frequency fc of the rotating machine 4, or may be a part of that range. The range of the rotational frequency fc may be calculated based on the current frequency fm of the electric motor 3 and the pulley ratio (R1:R2) according to Equation 1.

Here, comparing (A) and (B) in FIG. 6, it can be seen that the amplitude value a2 of the spectrum peak 602 in the abnormal state of (B) is smaller than the amplitude value a1 of the spectrum peak 601 in the normal state of (A). This is due to the following action. That is, if there is an abnormality in the transmission of the belt 5, the driving force corresponding to the load on the electric motor 3 side is not transmitted properly and efficiently to the rotating machine 4 side. As a result, in the abnormal state, a rotation frequency other than the rotation frequency fc, which is the rated frequency observed in the normal state, appears, and a sideband wave 603 in a frequency band including that rotation frequency is observed. As a result, the amplitude value a2 of the spectrum peak 602 is smaller than the amplitude value a1 of the spectrum peak 601. The amplitude values a1 and a2 are values in the range from 0 to 1.

In step S109, the diagnostic unit 13 stores the data of the detected spectrum peak value P for each accumulation time TL (including frequency and amplitude information as shown in FIG. 6) in memory (FIG. 2). Note that while the example in FIG. 6 shows a case where a peak appears at one frequency, the rotation frequency fc, this is not limited to the above, and in some actual cases, peaks may appear at multiple frequencies. Even in such cases, the diagnostic process can be applied in the same way.

[Process flow (5)-Status diagnosis process]
Next, in step S110, the diagnosis unit 13 performs a calculation process of the first information S1 representing the belt transmission abnormality degree based on the data of the above-mentioned spectrum peak value P. In this example, the diagnosis unit 13 compares the spectrum peak value P (frequency and amplitude) with a set threshold value H (i.e., spectrum peak threshold value) to determine the state value of the first information S1. The threshold value H is a threshold value related to the determination of the first information S1 representing the belt transmission abnormality degree, and particularly includes the amplitude threshold value. In this example, the diagnosis unit 13 determines the value of the first information S1 to be 0 when the spectrum peak value P is less than the threshold value H (P<H), and determines the value of the first information S1 to be 1 when the spectrum peak value P is equal to or greater than the threshold value H (P≧H). In the definition of the state in the first embodiment, the first information S1 takes a value of 0 or 1. S1=0 is a normal state, and S1=1 is an abnormal state. The closer the spectral peak value P is to the threshold value H, the closer the value of the first information S1 is to 0 (normal), indicating a state with a small degree of abnormality, and the farther the spectral peak value P is from the threshold value H, the closer the value of the first information S1 is to 1 (abnormal), indicating a state with a large degree of abnormality.

Figure 6 (C) shows an explanatory diagram regarding the above judgment. The amplitude value of a certain spectral peak value P is ap. The amplitude threshold value is ah. If the amplitude value ap of the spectral peak value P is equal to or greater than the threshold value ah, the value of the first information S1 is set to 0, and if the amplitude value ap of the spectral peak value P is less than the threshold value ah, the value is set to 1.

Next, in step S111, the diagnostic unit 13 uses the first information S1 to determine the value of the second information S2 related to the diagnosis result and notification, and determines the output control content according to that value. In the definition of the second information S2 in this example, the second information S2 takes the value 0 or 1. S2=0 indicates that no notification will be made because the condition is normal, and S2=1 indicates that a warning regarding an abnormality will be issued. In this example, the diagnostic unit 13 sets S2=0 when S1=0, and sets S2=1 when S1=1.

In step S112, if S2=0 (N), the diagnosis unit 13 returns to step S102, and if S2=1 (Y), the diagnosis unit 13 proceeds to step S113. If the process returns to step S102, the process is repeated in the same manner. In step S113, the diagnosis unit 13 transmits a signal corresponding to S1=1 (abnormal) to the display device 14, thereby causing the display screen of the display device 14 to display an image indicating that the power transmission mechanism 6 is in an abnormal state. In addition, the diagnosis unit 13 transmits a signal corresponding to S2=1 to the alarm device 15, thereby causing the alarm device 15 to output an alarm to warn that the power transmission mechanism 6 is in an abnormal state.

In step S113, the diagnostic unit 13 sets the value of flag F to 0 to perform the diagnosis end process, thereby ending the flow.

[Status Determination]
In the first embodiment, as a method for determining the state of the power transmission mechanism 6 after the normalization process in step S107, a method of determining by comparing the spectrum peak value P with a certain threshold value H is used as described above. This example of the method is not limited to this, and various methods can be applied. As another example of the method, a method may be used in which a plurality of threshold values and corresponding plurality of ranges are used to classify the state value of the first information S1 into three or more state values. As another example of the method, a method may be used in which a time threshold is further used to determine that a state such as an abnormal state exists when a state of a certain parameter value continues for a time period equal to or longer than the time threshold.

[Screen example]
FIG. 7 shows an example of a screen displayed on the display device 14, which displays the state of the power transmission mechanism 6 of the rotating machine system 100. Each part 101 including the power transmission mechanism 6 in FIG. 1 is assigned with management identification information (ID). The state value of the first information S1 is displayed for each ID. The diagnostic device 10 displays an image representing the state value so that the user U1 can easily recognize the state of the power transmission mechanism 6. In the screen example of (A), when S1=0, a "normal" character string image is displayed. The diagnostic device 10 continues to display this "normal" state as long as S1=0. In the screen example of (B), when S1=1, an "abnormal" character string image is displayed. When the value of the first information S1 changes from 0 to 1, the diagnostic device 10 switches to displaying this "abnormal" state. Note that the screen is not limited to displaying such a state, and may output, for example, a message prompting the user U1 to take action such as maintenance work.

(C) of FIG. 7 shows an example of a display in the case where there are multiple parts 101 including the power transmission mechanism 6 to be diagnosed in the rotating machine system 100, and the state of each of the multiple parts 101 is diagnosed in parallel at the same time. Even in this case, this system allows for collective monitoring of the state of the multiple parts 101. In the example screen of (C), there are four power transmission mechanisms 6, and images of the first information S1 for each ID are displayed in parallel. For example, only the part 101 of the power transmission mechanism 6 with ID=003 is in an abnormal state with S1=1. The display communication 14 may be a desktop PC, or an information terminal such as a tablet or smartphone. In this case, communication may be performed from the diagnostic device 10 to the information terminal using a communication interface such as Wi-Fi (registered trademark) or Bluetooth (registered trademark).

[Output control example]
Other examples of output control may be as follows. When the value of the first information S1 is, for example, 1 (abnormal) as a diagnosis result, the diagnostic device 10 automatically controls the switch 2 in FIG. 1 to switch the switch 2 to the OFF state by giving a control signal to the switch 2. This stops the operation of the rotating machine 4 through the power transmission mechanism 6 from the electric motor 3. The diagnostic device 10 also outputs a message to the effect that the rotating machine 4 has been stopped on the display device 14 or the like, and prompts the user U1 to take action such as maintenance work. Similarly, control of a transition to deceleration operation is also possible.

[Other screen examples]
The diagnostic device 10 may also display graphs such as those shown in Figs. 4, 5, and 6 on the display screen of the display device 14. In particular, an envelope normalized signal Ist such as that shown in Fig. 5 may be displayed. Also, a spectrum such as that shown in Fig. 6 may be displayed together with values of frequency, amplitude, threshold H, and the like. Also, on the screen of the display device 14, a user can set the above-mentioned setting unit 17. The user setting allows the selection of a judgment method (for example, the method of each embodiment) and the setting of a judgment threshold. Also, the user setting allows the setting of a timer for performing diagnosis and the management of information on the target rotating machine 4, etc. Also, the user setting allows the setting of a range of fluctuation in the load of the electric motor 3 and the rotating machine 4.

[Effects etc. (1)]
As described above, according to the diagnostic device 10 and diagnostic method of the first embodiment, even when there is a fluctuation in the load of the electric motor 3 and the rotating machine 4, it is possible to diagnose with high accuracy an abnormality or other condition related to the power transmission mechanism 6, for example, an abnormality in the transmission of the rotational driving force by the belt 5, while suppressing or preventing erroneous diagnosis or the like. In particular, according to the first embodiment, a current sensor is all that is required at the very least, and there is no need to install additional sensors or the like that measure the load of the electric motor 3 or the rotating machine 4, so that diagnosis can be realized at low cost. According to the first embodiment, since it is possible to perform a diagnosis with high accuracy while suppressing erroneous diagnosis or the like, it is possible to reduce downtime costs due to regular maintenance, even for devices that operate continuously for long periods of time or have complex mechanisms.

In the case of the conventional state diagnosis technology for rotating machine systems, the method of diagnosing the state of the power transmission mechanism using a fixed threshold value cannot respond to changes in the load on the rotating machine, and there is a high possibility of erroneous judgments occurring. As a countermeasure against erroneous judgments, state diagnosis technology that sets threshold values that vary according to the load can be considered. However, with this technology, it is necessary to set appropriate threshold values according to each load, and it is also necessary to install additional sensors (e.g., rotation sensors) that measure the load on the rotating machine. Also, a threshold data table that sets the correspondence between the varying load and threshold values based on the load information from the sensors, etc., and memory resources to store the table are also required. In other words, with this technology, additional costs increase and memory data volume becomes congested.

According to the first embodiment, even if the load of the rotating machine 4 changes based on the detection of the current supplied to the electric motor 3 during operation of the rotating machine system, the same algorithm can be used to detect abnormalities in the power transmission mechanism 6. In this method, it is basically necessary to detect and process the current to the electric motor 3 as shown in FIG. 1, and there is no need to install special means such as a sensor to measure the load of the rotating machine 4. In addition, in this system, when the diagnostic device 10 detects an abnormal state, it can immediately issue a warning to inform the user U1 of the occurrence of an abnormality in the belt 5, etc. The user U1 may be a worker performing processing work. In addition, the output of the notification, etc. is not limited to an output in the vicinity of the diagnostic device 10, but may also be a notification to a remote predetermined location or destination via communication.

[Effects etc. (2)]
The effects of the first embodiment will be described in more detail with reference to Figures 8 to 10. As described above, in this system, the rotational frequency fc on the rotating machine 4 side can be extracted from the processing results of the current information on the electric motor 3 side, and the state of the power transmission mechanism 6 can be determined using this information. In particular, in the first embodiment, the rotational frequency fc can be extracted from a spectrum peak that has undergone envelope normalization processing. By using such data, even when there is a load fluctuation, a highly accurate state determination is possible using an algorithm such as a constant threshold value H.

In the rotating machine 4 in FIG. 1 where load fluctuations occur, we will explain the change in the spectrum peak value P obtained in step S109 in a configuration in which normalization processing is performed in step S107 in FIG. 3 as in embodiment 1, and in a comparative example in which the normalization processing is not performed.

Figure 8 (A) shows an example of load fluctuation in a rotating machine system (part 101 in Figure 1) where the load on the rotating machine 4 increases monotonically over time. In the graph in (A), the horizontal axis shows time, and the vertical axis shows the load on the rotating machine 4 [dimensionless quantity]. An increase in the load on the rotating machine 4 corresponds to an increase in the rotational work of the electric motor 3.

Figure 8 (B) shows an example of the current detected by the electrical information detection unit 11 when the load on the rotating machine 4 increases as in (A). In the graph of (B), the vertical axis is the current [A]. As in (B), an increase in the rotational work of the electric motor 3 increases the load on the rotating machine 4, and it is known that the peak current value increases as the load on the rotating machine 4 increases. An example of the accumulation time TL is also shown in a dashed frame in (B).

Figure 9 (A) shows the results of the current in Figure 8 (B) when data is repeatedly collected for each accumulation time TL in step S105 and envelope detection processing is performed in step S106. In the graph in (A), the vertical axis is the current [A] of the envelope signal IhL. In this way, when the load fluctuates, the current value after envelope detection increases along with the load.

Figure 9 (B) shows the spectral peak value P when steps S108 and S109 are performed on the current value after envelope detection in (A). The spectral peak value P in the graph in (B) is shown as a transition on the time axis of the tip portion of the spectral peak represented by frequency and amplitude as in Figure 6. Line 901 shows the transition of the spectral peak value P when the power transmission mechanism 6 is in a normal state. Line 902 shows the transition of the spectral peak value P when it is in an abnormal state. The graph also shows line 903, which represents a constant amplitude threshold as the aforementioned threshold H.

Comparing the normal state line 901 and the abnormal state line 902, it can be seen that the spectrum peak value P increases with an increase in load in both states. In this case, for example, in step S110, if the threshold value H for the state value of the first information S1 is set to a constant value independent of the load as shown in the line 903, there is a risk of the following. That is, depending on the load, even if the actual state is normal, the spectrum peak value P may fall below the threshold value H, for example, at time 911, and the state may be erroneously determined to be abnormal. Also, depending on the load, even if the actual state is abnormal, the spectrum peak value P may exceed the threshold value H, for example, at time 913, and the state may be erroneously determined to be normal. Also, if the threshold value H is configured to change depending on the load, as described above, it is necessary to install a sensor or the like that measures the load, which leads to increased costs, etc.

Next, using FIG. 10, a case where normalization processing in step S107 is performed in a rotating machine 4 where load fluctuations occur will be described. Here, the premise of the load in FIG. 8 increasing monotonically is the same. FIG. 10 (A) shows the result of envelope detection in step S106 of data for each accumulation time TL collected in step S105 for the current in FIG. 8 (B), and normalization in step S107. In the graph in (A), the horizontal axis is time, and the vertical axis is the current value of the envelope signal IhL. This current value has a value in the range from 0 to 1 due to the normalization effect. In this way, since the current information is normalized to a value within a predetermined range in advance in step S107, the data for each accumulation time TL (current value shown) will always be within the predetermined range even if the load fluctuates.

Here, the information on the current amplitude (e.g., current amplitude 802) shown in FIG. 8B is lost due to the normalization process in step S107.

Next, FIG. 10B shows the spectral peak value P when steps S108 and S109 are performed on the current data in (A). Line 1001 shows an example of the transition of the spectral peak value P in a normal state. Line 1002 shows an example of the transition of the spectral peak value P in an abnormal state. Line 1003 shows an example of a constant value (amplitude threshold) as the threshold H.

As shown in the figure, the spectrum peak value P has an amplitude that converges within the range of 0 to 1 under both normal and abnormal conditions. Unlike FIG. 9B, both lines 1001 and 1002 converge to a constant value. It can be seen that the normal state line 1001 is a higher value than the abnormal state line 1002 because the power generated by the electric motor 3 is efficiently transmitted through the power transmission mechanism 6. Furthermore, because normalization processing is performed in step S107, the spectrum peak does not increase as shown in the figure even when the load increases.

Therefore, in the diagnostic device 10 and diagnostic method of the first embodiment, the diagnostic process including the normalization process of step S107 in FIG. 3 makes it possible to diagnose the state of the power transmission mechanism 6 even when there is a load fluctuation in the rotating machine 4. That is, as shown in FIG. 10, the state of the power transmission mechanism 6 can be diagnosed by comparing the spectrum peak value P, which represents the above-mentioned rotational frequency fc, with a certain threshold value H.

In the first embodiment, the data and timing that are the subject of normalization processing are the data and timing after the envelope detection processing in step S106, and data processing is performed every predetermined time (accumulation time TL). This also has the effect of reducing the amount of processed data and processing time.

<Embodiment 2>
A diagnostic device and a diagnostic method according to a second embodiment of the present invention will be described with reference to Figures 11 to 13. The basic configuration of the second embodiment is the same as that of the first embodiment, and the following describes the components of the second embodiment that are different from the first embodiment. In the second embodiment, learning is used for the diagnostic process.

[Diagnosis unit - processing flow]
Fig. 11 shows a flow of diagnostic processing by diagnostic unit 13 in diagnostic device 10 of embodiment 2. Fig. 11 has steps S201 to S216. Steps S203 to S209 and S210 to S214 in the flow in Fig. 11 are similar to steps S103 to S109 and S110 to S114 in the flow in Fig. 3, and steps S201, S202, S215, and S216 are different.

In step S201, the diagnosis unit 13 sets flag F. In the second embodiment, flag F has a value of 0 indicating stop, a value of 1 indicating start, and a value of 2 indicating learning. In the second embodiment, learning is possible in the diagnosis device 10, and F=2 is the value when learning is performed (referred to as "learning mode"). In step S202, if the value of flag F is 1 or 2, the flow proceeds to step S203, and if the value of flag F is 0, the flow ends. In steps S203 to S209, the diagnosis unit 13 performs the same process as in the first embodiment, and obtains the spectrum peak value P as a result of step S209.

Next, in step S215, the diagnosis unit 13 checks whether the value of flag F is 2, i.e., whether the learning mode is active. If F=2, the process proceeds to step S216, and if F=1, the process proceeds to step S210.

In step S216, the diagnosis unit 13 performs normal data learning processing. The diagnosis unit 13 learns the spectrum peak value P of the calculation result in step S209 when the power transmission mechanism 6 is in a normal state as normal data. Data related to the learning is stored in memory. After step S216, the flow ends and similarly repeats. In step S216, in the example of embodiment 2, the diagnosis unit 13 saves the arithmetic mean value of the spectrum peak values P in memory as the spectrum peak Ph of normal data. This spectrum peak Ph of normal data is calculated by the following equation 7. Σ is the sum of the spectrum peak values P from time t=0 to n. Ph is the value obtained by dividing this sum by n.
Equation 7: Ph={ΣP(t)}/n

If learning has been completed, the learning data is referenced in step S210. In step S216, the diagnosis unit 13 references the spectrum peak Ph of the learned normal data and the spectrum peak value P obtained in step S209. The diagnosis unit 13 calculates a state value (referred to as S1(t)) of the first information S1 indicating the degree of transmission abnormality of the belt 5 from the spectrum peak Ph of the normal data and the spectrum peak value P. In this case, the state value S1(t) is a value for each time point t, and is calculated by the following formula 8. P(t) is the spectrum peak value P for each time point t.
Equation 8: S1(t)=1-P(t)/Ph

Next, in step S213, the diagnostic unit 13 compares the state value S1(t) of the first information S1 with a preset threshold value H. If the state value S1(t) is less than the threshold value H, the diagnostic unit 13 determines that the state is normal and sets the value of the second information S2 to 0. If the state value S1(t) is equal to or greater than the threshold value H, the diagnostic unit 13 determines that the state is abnormal and sets the value of the second information S2 to 1. In step S214, if S2=1, the diagnostic unit 13 issues a warning to inform the user of the abnormal state in step S213. In step S214, the flag F is set to 0, and the flow ends.

[Display example]
FIG. 12 shows an example of output to the screen of the display device 14 in the second embodiment. (A) shows an example of a screen when the current time is a normal state (S1<H). In this screen, an image showing the "normal" state at the current time and a graph related to the transition of the state on a time series up to the current time are displayed. In the graph, the horizontal axis is time (t) and the vertical axis is the state value of the first information S1 showing the degree of transmission abnormality, and a line showing the threshold value H is also displayed. (B) shows an example of a screen when the current time is an abnormal state (S1≧H). In this screen, an image showing the "abnormal" state at the current time and a similar graph are displayed. By looking at the screen as shown in FIG. 12, the user U1 can confirm the transition of the state value (degree of abnormality) of the first information S1. The diagnostic device 10 updates the screen display contents in real time according to changes in the state. A similar display can also be applied to the case of collective monitoring of multiple rotating machines 4, etc.

[Effects, etc.]
As described above, according to the second embodiment, in addition to the same effects as those of the first embodiment, it is possible to determine an abnormal state based on learning data of normal states, and to quantitatively present the degree of abnormality (state value S1) of the power transmission mechanism 6 to the user U1 in a time series. This makes it easier to carry out maintenance and inspection work such as replacing deteriorated parts.

<Third embodiment>
A third embodiment of the present invention will be described with reference to Fig. 13. The third embodiment is a modification of the second embodiment. In the second embodiment, the state value of the first information S1 is calculated from the spectral peak value P and the arithmetic mean value Ph of the spectral peak values P in step S210 of Fig. 11. In the third embodiment, in calculating the state value of the first information S1, the Mahalanobis distance is calculated as follows, and this Mahalanobis distance is set as the state value.

[Mahalanobis distance]
In order to calculate the Mahalanobis distance, the diagnosis unit 13 calculates the average value, standard deviation, and correlation matrix within the normal data in the above-mentioned step S216 to create a Mahalanobis space. When the spectrum peak value P of the normal data and the state quantity related to the spectrum peak value P are given by matrix X, it is shown in the following formula 9. From matrix X of formula 9, the Mahalanobis distance is calculated as in formulas 10 and 11. Let the Mahalanobis distance be represented by D. Xm represents the average value of the state quantity. σ represents the standard deviation of the state quantity. R represents the correlation matrix. Xm ^T represents the transposed matrix of Xm.
Formula 9: X=[X ₁ , X ₂ , ..., X _n ]
Equation 10: Xm=[(X ₁ -X _m1 )/σ ₁ , (X ₂ -X _m2 )/σ ₁ , ..., (X _n -X _mn )/σ _n ]
Equation 11: D=√{1/n×XmR ⁻¹ ×Xm ^T } (D≧0)

The processing flow of the diagnostic device 10 in the third embodiment is the same as that shown in FIG. 11. The Mahalanobis distance D in the case where normal data has been learned in step S216 for a certain period of time (when flag F=0 has been set after a certain period of time has elapsed since flag F=2 was input in step S201) will be described below.

Figure 13 is an explanatory diagram of the Mahalanobis distance D. (A) in Figure 13 shows time series data of the spectral peak value P from the time ta when flag F=2 is input to the time tb when F=0 (i.e., the learning time). In this graph, the horizontal axis is time (t) and the vertical axis is the spectral peak value P. The upper point group shows normal data 1301 in a normal state, and the lower point group shows abnormal data 1302 in an abnormal state. The abnormal data 1302 has a reduced spectral peak value P compared to the normal data 1301. Here, the Mahalanobis distance D is calculated for data 1303 shown by black dots in (A).

(B) of FIG. 13 shows the probability density function of normal data 1311 and abnormal data 1312 with respect to the value obtained by dividing the spectrum peak value P by the standard deviation σ (referred to as the "normalized distance"). In this graph, the horizontal axis is the normalized distance (P/σ) [dimensionless quantity], and the vertical axis is the probability density [dimensionless quantity]. The normal data 1311 corresponds to the normal data 1301, and the abnormal data 1312 corresponds to the abnormal data 1302. The average 1313 is the average of the normal data 1311. Based on the above-mentioned formulas 10 and 11, the Mahalanobis distance D can be defined as the normalized distance (P/σ) from the average 1313 of the normal data 1311 to the target data 1303, as shown in the figure. The Mahalanobis distance D for each data can be used as a state value representing the degree of transmission abnormality. In other words, the larger the Mahalanobis distance D, the greater the degree of abnormality.

[Effects, etc.]
According to the third embodiment, in addition to the same effects as those of the second embodiment, the following effects are obtained. In the case of a rotating machine 4 with a load fluctuation, for example, a rotating machine 4 with large data variance of the spectrum peak value P, if there is still a small amount of learning data, the variation in the state value of the first information S1 also becomes large. By increasing the amount of learning data, the standardized distance and the Mahalanobis distance D are updated. This makes it easier to distinguish between normal data and abnormal data, as in the example of FIG. 13, i.e., the accuracy of diagnosis is improved.

In the example of Figure 13, a one-dimensional Mahalanobis distance using one variable, the spectrum peak value P, and in particular the arithmetic mean value Ph, was described. However, if various state quantities can be acquired by a sensor or the like, a multidimensional Mahalanobis distance using multiple variables can be calculated in the same manner and used as the first information S1.

<Fourth embodiment>
A fourth embodiment of the present invention will be described with reference to Fig. 14. The fourth embodiment is a modified example of the first embodiment. Fig. 14 shows the configuration of a rotating machine system 100 including a diagnostic device 10 of the fourth embodiment. In the fourth embodiment, the above-mentioned diagnostic device 10 is implemented so as to be built into the switch 2 and the electric motor 3 as an integral part. In the fourth embodiment, in particular, the diagnostic device 10 is implemented as a part of a controller 31 of the electric motor 3.

In FIG. 14, the electric motor 3 has a switch unit 32, a controller 31, and a motor unit 33. The function of the switch unit 32 is the same as that of the switch 2 described above. The switch unit 32 and the controller 31 may be implemented as one unit. The motor unit 33 is a mechanical structure such as a rotor and a stator that constitute a three-phase induction motor, in other words, a drive unit. The controller 31 includes a drive control unit 34 that is a part that controls the drive of the motor unit 33, and can be composed of, for example, a power conversion device such as an inverter, an IC board, etc. The controller 31 also has the electrical information acquisition unit 11 and diagnostic unit 13 implemented therein. A configuration in which the electrical information acquisition unit 11 and diagnostic unit 13 are implemented in the switch unit 32 is also possible.

As a result of the diagnostic process by the diagnostic unit 13, the first information S1 and second information S2 described above are output. These pieces of output information are output from the controller 31 via communication to the external display device 14 and notification device 15. The diagnostic unit 13 may also output an instruction for output control according to the first information S1 and the second information S2, such as an instruction for stopping operation or controlling deceleration operation, to the drive control unit 34. The drive control unit 34 drives and controls the motor unit 33 according to the instruction. The user U1 uses the functions of the diagnostic device 10 by operating the operation unit 16 or the like external to the controller 31.

As described above, according to the fourth embodiment, the condition of the power transmission mechanism 6 can be diagnosed using the diagnostic device 10 that is integrally implemented in the electric motor 3, and drive control can also be performed to deal with abnormalities, etc., depending on the results of the condition diagnosis.

<Fifth embodiment>
A fifth embodiment of the present invention will be described with reference to Fig. 15. The fifth embodiment is a modified example of the first and fourth embodiments. Fig. 15 shows a configuration of a rotating machine system 100 including a diagnostic device 10 of the fifth embodiment. In the fifth embodiment, the diagnostic device 10 described above is implemented by a communication device 50 and a diagnostic server 500. The communication device 50 is connected to the part 101. The diagnostic server 500 is disposed remotely from the communication device 50 via a communication network 501. The communication network 501 is a wide area network, a LAN, or the like.

The communication device 50 is disposed, for example, in close proximity to the part 101 including the electric motor 3 and the power transmission mechanism 6. The communication device 50 has an electrical information detection unit 51, an electrical information storage unit 52, and a communication unit 53. The electrical information detection unit 51 is connected to the power line 1L through a detection wire 1S, for example, as in the first embodiment. The electrical information detection unit 51 detects current, as in the electrical information detection unit 11 of the first embodiment. The electrical information detection unit 51 also includes an ADC that converts an analog current signal into a digital signal, etc.

The electrical information storage unit 52 stores and holds the detected current data, which is the electrical information obtained by the electrical information detection unit 51, in a memory consisting of a non-volatile memory or the like. The communication unit 53 includes a communication interface device, and transmits the detected current data stored in the electrical information storage unit 52 to the diagnostic server 500 via the communication network 501. Wi-Fi (registered trademark), Bluetooth (registered trademark), or the like can be used as the communication interface. This data communication may be performed automatically at a preset timing, or may be performed in response to a request from the diagnostic server 500.

The diagnostic server 500 includes a communication unit 54, an electrical information acquisition unit 55, a database (DB) 56, and a diagnostic unit 13. The communication unit 54 includes a communication interface device and performs communication processing with the communication device 50 via a communication network 501. The communication unit 54 acquires detected current data from the communication device 50 and stores it in a memory or DB 56. The diagnostic unit 13 performs diagnostic processing using the acquired data in the same manner as in the first embodiment, and stores the results in the DB 56 and outputs them from the display device 14 or the alarm device 15. The user U1 can recognize the state of the power transmission mechanism 6 of the rotating machine system 100 by using the diagnostic server 500 based on operations from the operation unit 16 or his/her own PC, etc.

Similarly, the diagnostic server 500 may collect and diagnose data from multiple parts 101, not limited to a part 101 including one rotating machine 4, and store the data in the DB 56. The diagnostic server 500 may perform analysis processing, etc., using the information in the DB 56. The user U1 can display and check the information in the DB 56 on the screen of the display device 14 as appropriate.

As described above, according to the fifth embodiment, electrical information can be temporarily stored in the communication device 50, and the condition can be diagnosed by the remote diagnostic server 500. Since electrical information can be stored in the communication device 50, there is no need to reduce the detection frequency of the electrical information detection unit 51, even if the communication speed of the communication network 501 is slow, for example. Furthermore, since the communication device 50 including the electrical information detection unit 51 and the diagnostic server 500 including the diagnostic unit 13 can be completely separated, for example, a user U1 such as a maintenance inspector can remotely diagnose the power transmission mechanism 6, making it possible to reduce the risk of entanglement accidents during inspection and the amount of inspection work required.

<Sixth embodiment>
A sixth embodiment of the present invention will be described with reference to Figures 16 and 17. In the first embodiment and the like, a method for diagnosing the state of the power transmission mechanism 6 based on the relationship between the current frequency fm of the electric motor 3 and the rotational frequency fc of the rotating machine 4, in particular, a state diagnosis method that can respond to load fluctuations by using normalization processing, is shown. In contrast, the sixth embodiment shows a state diagnosis method that can respond to load fluctuations without using normalization processing, based on the relationship between the current frequency fm of the electric motor 3 and the rotational frequency fc of the rotating machine 4. In the sixth embodiment, the rotational frequency fc, which is rotation information on the rotating machine 4 side, is obtained as an estimated rotational frequency by an estimation calculation.

Figure 16 shows the configuration of a rotating machine system 100 including a diagnostic device 10 of embodiment 6. In addition to the components of Figure 1, the diagnostic device 10 has a rotation information calculation unit 12. The rotation information calculation unit 12 inputs a current signal E1 (or digitized detected current data), which is electrical information, from the electrical information detection unit 11, and calculates rotation information on the rotating machine 4 side. The rotation information calculation unit 12 outputs processing result information including the rotation information to the diagnostic unit 13. The diagnostic unit 13 uses the information to perform diagnostic processing, the details of which differ from those of embodiment 1 (Figure 3).

The diagnostic device 10 of the sixth embodiment obtains an estimated rotational speed (Efm) of the electric motor 3 and an estimated rotational frequency (Efc) of the rotating machine 4 by performing estimation calculations. The estimated rotational speed Efm is estimated information corresponding to the current frequency fm. The estimated rotational frequency Efc is estimated information corresponding to the rotational frequency fc. The diagnostic unit 13 then diagnoses the state of the power transmission mechanism 6 based on the relationship between the estimated rotational speed Efm and the estimated rotational speed Efm.

As described above, when the transmission of the belt 5 of the power transmission mechanism 6 is in a normal state, the relationship between the current frequency fm of the electric motor 3 and the rotation frequency fc of the rotating machine 4 is expressed by Equation 1. The current frequency fm of the electric motor 3 is also a rotation frequency corresponding to the rotation speed of the electric motor 3, and can be known in advance by the frequency of the AC power supplied to the electric motor 3 or a sensor. In the first embodiment, the diagnostic device 10 detects a peak of a frequency (rotation frequency fc of the rotating machine 4 side) different from the peak of the current frequency fm of the electric motor 3, monitors or observes the change in the magnitude of the peak, and judges the state. The spectrum peak value P in the above-mentioned FIG. 6 shows the rotation frequency fc of the rotating machine 4. This makes it possible to diagnose and detect the presence of some abnormality or symptom in the belt 5, pulleys (3P, 4P), electric motor 3, or rotating equipment 4, which are representative components of the rotating machine system.

However, when the peak of the current frequency fm increases or decreases due to a change in the load on the rotating machine 4 (FIGS. 8 and 9), the peak of a frequency (say fx) different from the current frequency fm often increases or decreases following the peak of the current frequency fm. In the case of a technology that monitors only the peak of the different frequency fx in such a state where frequencies follow each other, even if the rotating machine system is operating normally, the peak of the different frequency fx increases, and this may result in a false determination that the state is abnormal or a symptom of an abnormality.

In addition, in the case of a technology that monitors only the magnitude of the peak of the different frequency fx, depending on the load connected to the rotating machine 4, the peak of the current frequency fm may decrease, but the peak of the different frequency fx may not change. This is the case when no abnormality occurs in the belt 5 or pulleys (3P, 4P) and the rotational driving force is properly transmitted from the electric motor 3 to the rotating machine 4. In this case, even though the rotating machine system is operating normally, the peak of the different frequency fx increases, and so it may be erroneously determined to be an abnormal or symptomatic state.

Therefore, in the diagnostic device 10 of the sixth embodiment, the current signal E1, which is the electrical information detected by the electrical information detection unit 11, is used to determine and detect the abnormality or other condition of the part 101 using a processing method that does not use the envelope detection of step S106 and the normalization process of step S107 in FIG. 3. As a result, in the sixth embodiment, the problem can be solved using a different processing method while still using the relationship between the current frequency fm of the electric motor 3 and the rotational frequency fc of the rotating machine 4 as in the first embodiment.

In FIG. 16, the rotation information calculation unit 12 can be realized by software program processing or a hardware circuit, etc. The rotation information calculation unit 12 obtains electrical information by sampling the current signal E1, which is electrical information measured from the power line 1L by the electrical information detection unit 11, at a predetermined time. The rotation information calculation unit 12 calculates an estimated rotation speed Efm from the detected current data, which is information in which the electrical information is arranged in chronological order, as shown in a speech bubble in FIG. 16, as a calculation to specify the current frequency fm of the electric motor 3.

Next, the rotation information calculation unit 12 compares the peak of the current frequency fm (corresponding estimated rotation speed Efm) with the peak of a frequency fx different from the current frequency fm. The rotation information calculation unit 12 calculates the estimated rotation frequency Efc as information corresponding to the aforementioned rotation frequency fc as the different frequency fx.

In other words, in the method of embodiment 6, the peak of the current frequency fm (corresponding estimated rotation speed Efm) and the peak of the different frequency fx (corresponding estimated rotation frequency Efc) are compared in relative value. Therefore, the information corresponding to the different frequency fx can be compared using the information corresponding to the current frequency fm as a reference.

The peak of the current frequency fm and the peak of the different frequency fx used in this comparison are not limited to the above-mentioned spectrum peak value P, but may be values obtained by smoothing the frequencies around these frequencies (fm, fx).

As a result, in the sixth embodiment, unlike the technology that monitors only the different frequencies fx (e.g., rotation frequency fc), even if the peak of the current frequency fm changes depending on the magnitude of the load as described above, the peak of the different frequencies fx is compared based on the magnitude of the peak of the current frequency fm. Therefore, it is possible to detect conditions such as abnormalities and symptoms in response to load fluctuations in the rotating machine 4.

As a supplement to the sixth embodiment, FIG. 17 shows an example of the peak of the current frequency fm and the peak of the different frequency fx according to the magnitude of the load. (A) is an example of a case where the load is small, and the spectrum peak 1701 has a current frequency fm and an amplitude value a11. The spectrum peak 1702 has a frequency fx different from the current frequency fm that follows the current frequency fm and an amplitude value a12. (B) is an example of a case where the load is large, and the spectrum peak 1703 has a current frequency fm and an amplitude value a13 (a13>a11). The spectrum peak 1704 has a frequency fx different from the current frequency fm that follows the current frequency fm and an amplitude value a14 (a14>a12). In (A), the relative value 1711 is the difference in amplitude between the spectrum peak 1701 and the spectrum peak 1702. In (B), relative value 1712 is the difference in amplitude between spectrum peak 1703 and spectrum peak 1704. The diagnostic device 10 uses the relative value of such a comparison between peaks to determine the state of the power transmission mechanism 6, for example, by comparing it with a threshold value.

[Effects, etc.]
As described above, according to the sixth embodiment, it is possible to obtain the same effects as those of the first embodiment, etc. Furthermore, according to the sixth embodiment, since the envelope detection process is not used, the amount of calculation can be reduced accordingly, and it becomes easier to implement the sixth embodiment in, for example, a microcomputer used in an embedded device.

In addition, in the sixth embodiment, in order to further improve the diagnostic accuracy, it is also possible to perform FFT processing on the acquired detection current data and then compare the peak of the current frequency fm with the peak of the different frequency fx.

Furthermore, in the sixth embodiment, a statistical process such as moving average or weighted average may be performed on the peak of the current frequency fm and the peak of the different frequency fx. In this case, the detection accuracy of an abnormality or other condition is improved in response to a time delay relative to the change in the load of the rotating machine 4. For example, the diagnostic device 10 calculates the average of the sampling value of the latest peak of the current frequency fm and the sampling value of the peak of the current frequency fm before the previous time. The diagnostic device 10 compares this average with the sampling value of the latest peak of the different frequency fx. Alternatively, the diagnostic device 10 may compare this average with the average of the latest peak of the different frequency fx and the peak of the different frequency fx before the previous time. These configurations are effective when an instantaneous shock is applied to the rotating machine system, or when electrical noise occurs around the AD conversion or in the power line 1L.

An example of envelope processing used in embodiment 6 is as follows. The diagnostic device 10 performs envelope processing on the detected current data obtained by plotting the current measured by the electrical information detection unit 11 in a time series. This envelope processing makes it possible to draw a curve connecting each peak in a composite wave having multiple frequency bands. Then, in the waveform after this envelope processing, it becomes possible to observe not only the sideband waves themselves generated by each frequency or signal of the composite wave, but also the overall picture of the electrical information. Note that the method of embodiment 6 can also be applied to the learning mode function of embodiment 2 in the same way.

[Modification]
The sixth embodiment is a configuration example in which the envelope detection process is not performed, but as a modified example of the sixth embodiment, a configuration example in which the envelope detection process is used will be described. As shown in the envelope signal IhL in (A) of FIG. 5 described above, in the waveform in which the process of detecting or acquiring the envelope has been performed, the part that repeatedly becomes convex with respect to time corresponds to the rotation frequency fc. The diagnostic device 10 can acquire the peak of the rotation frequency fc by measuring the peak in a predetermined section that takes the rotation frequency fc into consideration by a process such as moving average. On the other hand, the frequency fx different from the rotation frequency fc in the predetermined section is set to the waveform size other than the convex part of the rotation frequency fc. The waveform of this different frequency fx can be obtained by a process of integrating the waveform other than the convex part or a process of moving average the size of a predetermined timing.

By the above process, the diagnostic device 10 can obtain information on the waveform of a frequency fx that is different from the waveform of the rotational frequency fc. The diagnostic device 10 then compares the waveform of the different frequency fx with the waveform of the rotational frequency fc as a reference. In other words, the diagnostic device 10 looks at the relative value between the waveform of the rotational frequency fc and the waveform of the different frequency fx. As a result, even in this modified example, it is possible to detect conditions such as abnormalities and symptoms in response to load fluctuations in the rotating machine 4.

Although the present invention has been specifically described above based on the embodiments, the present invention is not limited to the above-mentioned embodiments and can be modified in various ways without departing from the scope of the present invention. A combination of the various embodiments is also possible. The technology disclosed herein is widely applicable to rotating machine systems including a power transmission mechanism. In each of the above-mentioned embodiments, the diagnostic device 10 obtains rotation information as information corresponding to the rotation frequency fc of the rotating machine 4 in relation to the current frequency fm of the motor 3 based on detection and measurement of current from the AC power source 1 to the motor 3. Then, by using this information, the diagnostic device 10 can diagnose the state of the power transmission mechanism 6 with a predetermined algorithm even when there is a load fluctuation in the rotating machine 4. When obtaining the above-mentioned rotation information, as described above, normalization processing, envelope detection processing, estimation calculation, frequency change (different frequency fx), etc. can be used. Not limited to this, any processing for obtaining the above-mentioned rotation information can be applied. For example, a method of counting the occurrence time of peak intervals in an envelope as shown in FIG. 5 and calculating the estimated rotation frequency Efc of the rotating machine 4 from the occurrence time of the peak interval can also be applied.

1...AC power source, 2...switch, 3...electric motor, 4...rotating machine, 3P...first pulley, 4P...second pulley, 5...belt, 6...power transmission mechanism, 10...diagnosis device, 11...electrical information detection unit, 13...diagnosis unit, 14...display device, 15...alarm device, 100...rotating machine system, 101...part.

Claims (7)

an electric motor driven by power supplied from a power source;
a power transmission mechanism connected to the electric motor;
a rotating machine connected to the power transmission mechanism; and
a diagnostic device connected to the electric motor or to a power supply line from the power source to the electric motor, for diagnosing a state of the power transmission mechanism;
The diagnostic device in a rotating machine system comprising:
the power transmission mechanism includes a first pulley that outputs a rotational driving force of the electric motor, a second pulley that drives the rotating machine, and a drive transmission component that connects the first pulley and the second pulley,
When a load applied to the rotating machine from the electric motor through the power transmission mechanism fluctuates,
The diagnostic device comprises:
Detecting a current in the power supply;
calculating an estimated rotation speed of the motor based on the detected current value in a time series;
Calculating an estimated rotational frequency of the rotating machine based on the estimated rotational speed of the electric motor;
extracting a peak of a current frequency corresponding to an estimated rotation speed of the electric motor and a peak of a frequency different from the current frequency corresponding to an estimated rotation frequency of the rotating machine;
a state of the power transmission mechanism is determined using a relative value obtained by comparing a peak of the current frequency with a peak of the frequency different from the current frequency based on a relationship between a current frequency corresponding to the estimated rotation speed of the electric motor and a frequency different from the current frequency corresponding to the estimated rotation frequency of the rotating machine, and the state of the power transmission mechanism is determined and a diagnosis result is stored and output.
Diagnostic equipment for rotating machinery systems. 2. The diagnostic device for a rotating machine system according to claim 1,
the diagnostic device calculates a spectral peak based on a spectral analysis for a frequency band including an estimated rotation frequency of the rotating machine, and determines a state of the power transmission mechanism using a frequency and an amplitude of the spectral peak.
Diagnostic equipment for rotating machinery systems. 3. The diagnostic device for a rotating machine system according to claim 2,
The diagnostic device comprises:
With learning mode,
In the learning mode, learning is performed using the spectrum peak when the power transmission mechanism is in a normal state as normal data , and data related to the learning including the normal data is stored in a memory;
After the learning is completed, when a diagnosis is performed other than in the learning mode, the state of the power transmission mechanism is determined based on the spectrum peaks in the normal data and the spectrum peaks calculated based on the spectrum analysis of the diagnosis target, using the data related to the learning .
Diagnostic equipment for rotating machinery systems. 4. The diagnostic device for a rotating machine system according to claim 3,
the diagnostic device determines a state of the power transmission mechanism based on a ratio between the spectrum peak in the normal data serving as a reference and the spectrum peak in the diagnostic target .
Diagnostic equipment for rotating machinery systems. 3. The diagnostic device for a rotating machine system according to claim 2,
the diagnostic device calculates a Mahalanobis distance using time series data of the spectrum peaks, and judges a state of the power transmission mechanism using the Mahalanobis distance.
Diagnostic equipment for rotating machinery systems. 3. The diagnostic device for a rotating machine system according to claim 2,
The diagnostic device comprises:
a communication device connected to the electric motor;
a server device connected to the communication device via a communication network;
having
The communication device stores data of the detected current and transmits the data to the server device;
the server device determines a state of the power transmission mechanism by using the data acquired from the communication device;
Diagnostic equipment for rotating machinery systems. an electric motor driven by power supplied from a power source;
a power transmission mechanism connected to the electric motor;
a rotating machine connected to the power transmission mechanism; and
a diagnostic device connected to the electric motor or to a power supply line from the power source to the electric motor, for diagnosing a state of the power transmission mechanism;
A diagnostic method for a rotating machine system comprising:
the power transmission mechanism includes a first pulley that outputs a rotational driving force of the electric motor, a second pulley that drives the rotating machine, and a drive transmission component that connects the first pulley and the second pulley,
When a load applied to the rotating machine from the electric motor through the power transmission mechanism fluctuates,
The steps executed by the diagnostic device include:
detecting a current in the power supply;
calculating an estimated rotation speed of the motor based on the detected current values on a time series basis;
calculating an estimated rotational frequency of the rotating machine based on the estimated rotational speed of the electric motor;
extracting a peak of a current frequency corresponding to an estimated rotational speed of the electric motor and a peak of a frequency different from the current frequency corresponding to an estimated rotational frequency of the rotating machine;
a step of determining a state of the power transmission mechanism using a relative value obtained by comparing a peak of the current frequency with a peak of a frequency different from the current frequency based on a relationship between a current frequency corresponding to an estimated rotational speed of the electric motor and a frequency different from the current frequency corresponding to an estimated rotational frequency of the rotating machine, and storing and outputting a diagnosis result;
A method for diagnosing a rotating machine system comprising:

Images