JP3294766B2 - Abnormality diagnosis device and method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an apparatus and method for diagnosing abnormality of equipment, and more particularly, to an abnormality of equipment capable of easily diagnosing an abnormality of equipment such as a rotating machine installed in an arbitrary equipment space. The present invention relates to a diagnostic device and a method.

[0002]

2. Description of the Related Art Conventionally, as a method for diagnosing abnormalities of equipment installed in an arbitrary equipment space, for example, a ceiling portion of a clean room, for example, a rotating machine such as a motor of a fan filter unit, a predetermined sensor (for example, vibration Sensors,
Temperature sensors, acoustic sensors, etc.)
There is known a method of diagnosing abnormality of equipment by measuring vibration, temperature, or sound of the equipment and determining whether or not the measured value deviates from a preset threshold. .

[0003] Further, there is also a method in which an operator skilled in equipment maintenance directly touches equipment such as a rotating machine without using the above-mentioned automatic diagnostic apparatus, and empirically judges abnormality of the equipment based on the tactile sensation. Well done.

[0004]

However, the conventional equipment abnormality diagnosis apparatus or method as described above has the following problems, for example. 1. In general, the mounting position of various sensors is limited by design, so that it is not always possible to install it at the optimum position where accurate measurement values can be obtained, and in some cases, the mounting of the sensor itself may not be possible . 2. When the number of objects to be monitored is large, such as a fan filter unit installed on the ceiling of a clean room, installing sensors on each equipment will increase the initial cost, and equipment installed adjacent to the equipment , Accurate values for each individual device cannot be obtained. 3. In the case of a known automatic diagnosis system, the initial value of each equipment is often unknown, and its identification is also difficult. In particular, when the number of monitored objects is large, the difficulty increases dramatically as the number of monitored objects increases. Further, even with a system having a large number of management points, it is difficult to easily follow the addition of equipment and the like. Also, when an attempt is made to introduce a known automatic diagnosis system into an existing building, the initial value is unknown. 4. In addition to manual diagnostic systems, even in the case of an automatic diagnostic system, it requires skill to determine whether a device is normal or abnormal based on measured values, and individual differences are likely to occur. Diagnosis is difficult. 5. In a conventional diagnostic system, not only in the case of a manual diagnostic system but also in the case of an automatic diagnostic system, the number of processing steps is relatively large and the amount of work of an operator is large.

[0005] Accordingly, the present invention has been made in view of the above-mentioned problems, and is capable of diagnosing an abnormality in equipment without requiring special training or experience and setting initial values by numerical values. Can be performed easily and reliably without individual differences, and even when many equipments are installed, simultaneous measurement of multiple equipments with a small number of sensors is possible. It is an object of the present invention to provide a new and improved apparatus and method for diagnosing abnormality of equipment, which has a high degree of freedom regarding the installation location of the sensor and can easily narrow down abnormal equipment.

[0006]

According to a first aspect of the present invention, there is provided a system for collecting sound generated by one or more monitoring objects installed in an arbitrary equipment space. There is provided an abnormality diagnosis device that collects sound by a sound device and diagnoses an abnormality of the monitoring target based on the sound signal.
The abnormality diagnosis device includes a standardization unit that sets a fluctuation width of an audio signal sampled over a certain time interval as a reference value, at a certain time (t).
Of the sampled acoustic signal (X (t))
Radius is relative size (A (i)) with respect to reference value
In the polar coordinate system, after τ hours from a certain time (t),
The reference value of the pulled acoustic signal (X (t + τ))
To the declination (Θ (i), Φ (i))
By transforming at least one start line of the polar coordinate system
At least two dots (A
Plot (i), Θ (i)) (A (i), Φ (i))
Plotting means, and display means for visually displaying the dot pattern plotted by the plotting means, and diagnosing an abnormality of the monitored object in accordance with the dot pattern displayed by the display means. And If the display means is configured to display the normal dot pattern and the diagnostic dot pattern on the same plane (for example, on the same screen or on the same paper), the visual comparison work can be performed more easily. It can be done easily and efficiently.

[0007] The abnormality diagnosis apparatus according to claim 2 is
A standardizing means for setting a fluctuation width of an audio signal sampled over a certain time interval as a reference value, and a certain time
The acoustic signal (X) sampled at (t)
(T)) relative to the reference value (A)
A certain time in a polar coordinate system having a radius of (i))
The acoustic signal (X
The relative magnitude of (t + τ)) to the reference value is declination
(Θ (i), Φ (i)) gives the polar coordinates
At least one line-symmetric about at least one start line of the system
And two dots (A (i), Θ (i)) (A (i), Φ
Plotting means for plotting (i)), pattern recognizing means for recognizing the dot pattern plotted by the plotting means, and a dot pattern at the time of diagnosis recognized by the pattern recognizing means and a normal state obtained in advance. A comparison diagnosis that compares the dot pattern at the time of diagnosis and the dot pattern at the time of normality with a predetermined value or more and diagnoses that the monitoring target is abnormal. It is characterized by having means.

A polar coordinate system used for obtaining a dot pattern is divided into a plurality of symmetrical regions by a plurality of (n) start lines, and two dots (a (i), By plotting b (i)), for example, by plotting dots with respect to six symmetry axes (start lines), a snow crystal dot pattern is obtained, and the pattern is visually and pattern-recognized. Accordingly, it is possible to easily and reliably identify a normal pattern and an abnormal pattern without setting a numerical initial value.

Further, according to a second aspect of the present invention, sound generated by one or more monitored objects installed in an arbitrary equipment space is collected by a sound collection device, and the sound signal is added to the sound signal. An abnormality diagnosis method for diagnosing abnormality of the monitoring target based on the abnormality diagnosis method is provided. The abnormality diagnosis method includes the steps of setting as claimed in claim 3, the variation width of a sampled sound <br/> sound signal over a time interval as the reference value; Oh
At time (t), the sampled acoustic signal (X
(T)) relative to the reference value (A)
A certain time in a polar coordinate system having a radius of (i))
The acoustic signal (X
The relative magnitude of (t + τ)) to the reference value is declination
(Θ (i), Φ (i)) gives the polar coordinates
At least one line-symmetric about at least one start line of the system
And two dots (A (i), Θ (i)) (A (i), Φ
(I)) plotting ; and repeating the step of plotting dots in the polar coordinate system over a predetermined period of time, and diagnosing an abnormality of the monitoring target based on a dot pattern obtained as a result. It is characterized by becoming.

[0010] When diagnosing an abnormality of the monitored object by the above-mentioned abnormality diagnosing method, the obtained dot pattern is visually displayed as described in claim 4, and the previously acquired normal pattern is obtained. It is also possible to compare the dot pattern with the dot pattern at the time of diagnosis and visually diagnose the abnormality of the monitored object.
As described in the above, using an appropriate pattern recognition method, the obtained dot pattern is pattern-recognized, and the dot pattern at the time of normality and the dot pattern at the time of diagnosis obtained in advance are compared with each other to perform the diagnosis. When there is a deviation equal to or more than a predetermined value between the dot pattern at the time and the dot pattern at the normal time, the monitoring target may be diagnosed as abnormal.

Further, according to a third aspect of the present invention, sound generated by a plurality of monitoring objects installed in a large facility space is collected by a sound collecting device, and based on the sound signal. A method for identifying an abnormal operation monitoring object is provided. The method further comprises dividing the global facility space into a plurality of smaller facility spaces each including a plurality of objects to be monitored, and narrowing down the global facility space; In the equipment space, the sun
Setting the fluctuation range of the coupled acoustic signal as a reference value; and sampling at a certain time (t).
Relative to the reference value of the acoustic signal (X (t))
In a polar coordinate system whose radius is the size (A (i)),
Sound signal sampled τ hours after the time (t)
(X (t + τ)) relative to the reference value
Is converted to the argument (Θ (i), Φ (i)),
Axisymmetric about at least one start line in polar coordinate system
At least two dots (A (i), Θ (i)) (A
(I), Φ (i)) ; and repeating the step of plotting dots in the polar coordinate system for a predetermined period of time. Based on the resulting dot pattern, an object to be monitored for abnormal operation is determined. A step of specifying a small facility space included; and a step of specifying a monitoring target of an abnormal operation from a plurality of monitoring targets included in the small facility space specified in the step. It is characterized by.

[0012] In the above method, before specifying the monitoring target of the abnormal operation, a small facility space including the monitoring target of the abnormal operation is further reduced before specifying the monitoring target of the abnormal operation. The step of re-dividing and narrowing down the space and narrowing down the equipment space including the monitoring target of the abnormal operation may be repeated a required number of times.

In the methods according to the second and third aspects, the polar coordinate system used to obtain the dot pattern is divided into a plurality of symmetrical regions by a plurality of (n) starting lines, and each starting line is For example, by plotting two dots (a (i), b (i)) which are line-symmetrical with respect to, for example, by plotting dots with respect to six symmetry axes (start lines), A crystalline dot pattern is obtained, and a normal pattern and an abnormal pattern can be easily and reliably distinguished visually and by pattern recognition without setting a numerical initial value.

[0014]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, preferred embodiments of an abnormality diagnosis apparatus and method according to the present invention will be described in detail with reference to the accompanying drawings.

First, referring to FIG. 1, a first embodiment of the present invention will be described.
The configuration of the abnormality diagnosis apparatus 10 according to the embodiment will be described. Reference numeral 12 denotes a fan filter unit (FF) which is a monitoring target having a rotation mechanism such as a fan motor.
U), for example, equipment installed in an arbitrary equipment space such as a ceiling of a clean room (not shown). 14
Is a sound collection device such as a microphone that collects sound generated from the fan filter unit 12. In the illustrated example, the microphone 14 is disposed immediately above the portion 12a where the rotation mechanism of the fan filter unit is installed. However, as described later, according to the present embodiment, the microphone 14 collects sound. Since the obtained acoustic signal is normalized, the attenuation of the acoustic signal due to the distance between the monitoring target 12 and the microphone 14 is canceled. Therefore, when diagnosing the same equipment, it can be considered that the sound collection position by the microphone 14 has little effect on the shape of the dot pattern obtained according to the present embodiment. In addition, it is not necessary to strictly define the sound collection position, and it is not necessary to particularly amplify the collected sound signal.

The sound signal collected by the microphone 14 is measured (sampled) at a predetermined cycle by sampling means such as a sound level meter 16, converted into data by an A / D converter 18, and processed by a signal processing unit (SDP converter). ) 19
Sent to The signal processing unit 19 plots the sampled acoustic signal on a polar coordinate system as shown in FIG.
Get the dot pattern. In this manner, the dot-patterned sound signal is displayed on the screen of the display device 20 such as a CRT, or printed out on paper using a printing unit such as a printer. Then, as described later, the worker can easily distinguish the abnormal pattern from the normal pattern from the dot pattern visually displayed on the screen or on the paper.
Note that when visually displaying the dot pattern, FIG.
As shown in (1), if the dot pattern 20a in the normal state and the dot pattern 20b in the diagnosis are displayed on the same plane, the abnormal pattern can be more easily identified.

Next, with reference to FIG.
9, a procedure for plotting a dot pattern of a sampled acoustic signal on a polar coordinate system will be described.
Note that the dot pattern depiction method adopted in the present embodiment is known as one of chaotic graphic expressions that characterize local differences in an opposite direction, such as “targeted dot pattern (SDP: Symmetrized).
Dot Patterns) method. Note that SDP
The law is a well-established graphic method, and its contents are described in Clifford. A. I'm familiar with Pickover's "Computer Chaos Fractal", translated by Tokichiro Takahashi and Shozo Naito, published by Shiraagesha Publishing.
The detailed description thereof will be omitted by referring to the book.

The targeted dot pattern (SDP) is a dot pattern obtained by plotting the correlation between the value X (t) at time t of the time-series data and the value X (t + τ) at time t + τ in a polar coordinate system. The procedure for plotting the target dot pattern will be described with reference to FIG.
A time waveform as shown in FIG.
A set of data sampled at [S] (X (t) , X
(T + τ)) is plotted as (a1, a1 ′) so as to be axisymmetric with respect to the start line of the polar coordinate system as shown in FIG. 2B, and the start line is further rotated by the number of symmetry axes. Shinaga
Et al., Is intended to continue to plot a set of data for KakuHajimesen. For example, as shown in FIG. 3, when the number of symmetry axes is six, it is possible to divide the polar coordinates into six target areas by the starting line rotated by 60 degrees. A set of 12 dots (a1, a1 ') ~
(A6, a6 ') can be plotted in a polar coordinate system.

The declinations Θ (i) and Φ (i) that are line-symmetric with the radius A (i) in the polar coordinate system of the time-series data X (i) discretized at a constant sampling time Δt are as follows: Converted by expression.

[0020]

(Equation 1)

[0021] As above clear expression, RBI dot (A (i), Θ ( i)), (A (i), Φ (i)) is one
The interval L of the data of the set and the gain ζ are used as parameters. Therefore, in order to identify abnormalities by SDP from time-series data, it is necessary to optimize the interval L and the gain ζ of a set of data . The time width L and the gain ζ can be appropriately selected according to the frequency characteristics of the acoustic signal generated from the monitoring target. For example, when the monitoring target is a rotating machine such as an FFU, an acoustic signal generated therefrom has a periodic function: sin (2πf
t), for example, if the fundamental frequency f = 550 Hz and Δt = 1 / 20,000 s,
By setting the interval L between a set of data to 5 to 6 and 程度 to about 40 , the shape of the SDP in which the outside of the
And it is easy to visually identify.

As is clear from the above equation, the sampled sound signal has a fluctuation width (Xma) of the sound signal.
x-Xmin), the attenuation of the acoustic signal due to the distance between the monitored object and the microphone is offset. Therefore, when diagnosing the same equipment,
Since the sound collection position by the microphone can be considered to have little effect on the shape of the dot pattern obtained by the present embodiment, according to the present embodiment, it is not necessary to strictly define the sound collection position, Further, it is not necessary to particularly amplify the collected sound signal. As for the background noise, it is sufficient to recognize the tendency of the sound as described later, so that it is not necessary to take measures for that.

FIGS. 4 to 6 show a normal FFU and an abnormal FF.
An example in which SDP expression is performed on the sounds generated from U by the abnormality diagnosis device according to the present embodiment will be described. FIG. 4 shows a comparison between an FFU in which an allophone sound was recognized and a normal FFU, and FIGS. 5 and 6 show a comparison between an FFU in which an allophone sound was recognized and a normal FFU. An example is shown. In each case,
The data of the FFU in which abnormal noise is recognized are shown on the right side.

As can be seen from these charts, it is possible to determine the difference between a normal sound signal and an abnormal sound signal from a graph in which a sound signal sampled by a microphone is simply represented as a time waveform. Have difficulty.
On the other hand, according to the SDP expression, a normal sound signal is clearly plotted with respect to the axis of symmetry, whereas an abnormal sound signal has an ambiguous plot with respect to the axis of symmetry. In the example, it can be seen that the dot pattern is developed in a circular shape. As described above, according to the SDP expression by the abnormality diagnosis device according to the present embodiment, the abnormal sound signal is expressed by the ambiguity of the expression. Diagnosis can be made easily visually. At this time, if the normal dot pattern and the dot pattern at the time of diagnosis are simultaneously displayed on the same screen as shown in FIGS. 4 to 6, it is possible to more easily determine whether the pattern is normal or abnormal. .

In the examples shown in FIGS. 4 to 6, the cause of the ambiguity in the SDP expression of the abnormal sound signal is as follows.
It is considered that when an abnormality occurs in the monitoring target that is a rotating system, the multiple of the shaft rotation frequency and the vibration acceleration of the nonlinear component increase. In addition, as in the example shown in FIG. 4, it is considered that the reason why the circumferential contour occurs in the SDP expression of the abnormal sound signal is that irregular amplitude occurs in the amplitude of a certain periodic function.

Further, in the example shown in FIG. 1, an apparatus configuration for diagnosing an abnormality of one FFU is shown, but the present invention is not limited to this example. For example, as shown in FIG. 7, a large number of FFUs (U11 to U11)
Umn). However, each of these FFUs
If a microphone is individually installed or an abnormality diagnosis is individually performed, initial costs are increased and a great deal of labor is required. In this regard, according to the present invention, it is not necessary to individually diagnose abnormalities of individual FFUs, and it is possible to collectively diagnose a plurality of FFUs included in a certain range at the same time.

That is, according to the apparatus and method of the present invention, even if sounds from a plurality of FFUs are collected simultaneously,
The acoustic signal can be regarded as a superposition of a plurality of periodic functions, and a plurality of FFUs included in a certain range can be considered.
Can be considered to represent a unique SDP. Therefore, if any of the FFUs included in the range is abnormally operated, a change also appears in the superimposed sound signal, so that it is possible to determine whether the range includes an abnormal FFU. For example, in the illustrated example, a microphone 34 is movably mounted on a ceiling by a driving device 36, for example, nine FFUs.
(U11 to U13, U21 to U23, U31 to U33)
The abnormality diagnosis can be performed using a region including the as a diagnostic unit. The apparatus and method of the present invention can specify an area including an abnormal FFU, and further narrow down the area, thereby easily specifying the abnormal FFU. Alternatively, the same operation as described above can be achieved by previously installing an appropriate number of microphones at predetermined intervals on a wall or the like, providing the same number of channels on the computer side, and switching the channels as needed.

As described above, according to the present invention, even when a large number of objects to be monitored such as FFUs are installed in a large facility space such as a ceiling of a clean room, the large facility space can be reduced. , Each of which is divided into smaller equipment spaces each including a plurality of monitored objects, and the acoustic signal generated from each of the small equipment spaces is represented by the SDP, so that the equipment space including the abnormal monitored objects is easily included. Can be narrowed down. Needless to say, the narrowing of the equipment space including the abnormal monitoring target can be performed in multiple stages. And after a certain amount of narrowing down,
It is possible to specify a monitoring target that is operating abnormally from the specified small facility space.

Next, with reference to FIGS. 8 and 9, a description will be given of a diagnosis flow for actually specifying a small facility space including an abnormal FFU from a large facility space.

First, as shown in FIG.
A normal SDP storage process is performed. That is, after the microphone is installed in the equipment space of the predetermined small area (step S81), the measurement of the acoustic signal is started (step S8).
2), A / D conversion is performed on the sampled acoustic signal at dT intervals for T seconds (step S83), and a digital signal is obtained (step S84). This signal is displayed as an SDP by the method described above (step S85), and the result is stored as a normal SDP (step S8).
6).

Next, a diagnostic routine for the rotating device shown in FIG. 9 is started. First, the user moves to the first measurement location (step S91), and installs a microphone (step S9).
2). Then, measurement of the acoustic signal is started in the installation space (step S93), and the sampled acoustic signal is A / D-converted at dT intervals for T seconds (step S94) to obtain a digital signal (step S95).
This signal is displayed as an SDP by the method described above (step S96), and is compared with a previously stored normal SDP in the preprocessing shown in FIG. 8 (step S97). If there is a difference, it is determined that there is an abnormality in the measurement space (step S98), and if there is no difference, it is determined that the inside of the measurement space is normal (step S99). (Step S10
0), end the diagnosis.

In the routines shown in FIGS. 8 and 9, a case is described in which an abnormality is diagnosed in real time at the measurement site. However, as in the routine shown in FIG. It is also possible to configure so that the abnormality is diagnosed while performing the measurement and reproducing the acoustic signal recorded later.

Although the preferred embodiments of the abnormality diagnosis apparatus and method according to the present invention have been described with reference to the accompanying drawings, the present invention is not limited to the above examples. It is obvious for those skilled in the art that various changes and modifications can be made within the scope of the technical idea described in the scope of claims, and these naturally fall within the technical scope of the present invention. It is understood to belong.

For example, in the above description of the embodiment, an example is shown in which the abnormality diagnosis apparatus and method according to the present invention are applied to determine whether the operation of the monitored object is normal or abnormal. However, since it is empirically known that a specific SDP expression appears for a specific failure, the SDP expression for each failure cause and the SDP at the time of diagnosis are known by knowing in advance the characteristics of the SDP expression for each failure cause. By comparing the expressions with the expressions, it is also possible to configure to determine the cause of the failure.

In the description of the above embodiment, an example in which an SDP expression is developed on a screen of a display device such as a CRT and the operator visually diagnoses an abnormality has been described. , It is also possible to configure so as to automatically perform the abnormality diagnosis.

FIG. 11 shows an example of a method for comparing SDP expressions in normal and abnormal states by pattern recognition.
First, a normal SDP and a diagnostic SDP are prepared, and the coordinate positions of the SDP are developed on a matrix. Specifically, the coordinates of one axis of SDP are
For example, it is divided into a 1000 × 1000 matrix. Normal SDP matrix: NSD [1000 × 1
000] SDP matrix when abnormal: DSD [1000 × 1
000]

If the initial value of the matrix is set to 0 and there is a plot on the coordinate axis, the value of the matrix corresponding to the coordinate is input as 1, and the SDP plot is developed into a matrix. In this way, the presence / absence of plots for the normal SDP and the abnormal SDP is developed into a matrix.

Next, the normal SDP is subtracted from the abnormal SDP developed in the matrix to obtain a deviation matrix HSD. HSD [i, j] = DSD [i, j] −NSD [i, j] However, when DSD [i, j] −NSD [i, j] = − 1, it is regarded as 0. This is because it can be determined that an abnormal SDP pattern is formed by moving the plotting position at normal time to another coordinate. That is, the abnormal pattern is plotted on coordinates that are not in the normal pattern.

Next, the deviation matrix is reconverted to SDP to obtain the deviation SDP. Then, an abnormality can be identified by performing pattern matching of the screen processing on the shape of the deviation SDP and the reference SDP indicating the abnormality.

In the above description of the embodiment, a rotating system such as an FFU is taken as an example of the monitoring object, but the abnormality diagnosis apparatus and method according to the present invention are not limited to this example. For example, even for a motion system other than a rotary system such as a reciprocating engine or a pump, the acoustic signals generated therefrom each have a unique SDP expression. The present invention can also be applied to determine abnormalities of various motor systems.

[0041]

As described above, according to the present invention,
Since the difference between the localized acoustic signals can be expressed globally, it is possible to visually and easily diagnose the abnormality of the monitored object without requiring much specialized knowledge.
Further, according to the present invention, a result can be obtained by a simple operation of measuring an acoustic signal and drawing an SDP,
The processing required for diagnosis is very simple, and a simple abnormality diagnosis device can be constructed.

Further, in the present invention, as a method for diagnosing abnormality, a comparison between the time of diagnosis and the time of normal is performed.
Although a graphic pattern by P is required, a numerical limit value is not required, and there is no difficulty in acquiring the system.
Furthermore, if one piece of SDP data of devices related to diagnosis is prepared in advance, it can be generally used for each facility.

Further, according to the present invention, the measured sound signal is normalized when drawing the SDP, so that the attenuation of the sound signal depending on the distance between the sound source and the sound collecting means is canceled. Therefore, when diagnosing the same device, the SDP shape is hardly affected by the distance between the sound collection device and the sound generation source, so that diagnosis can be performed easily and accurately.

Further, according to the present invention, even when a large number of objects to be monitored are present in a wide area, an acoustic signal generated from a predetermined collective space is represented by SDP, so that abnormalities in the space can be detected. Since it is possible to determine whether or not there is, it is possible to drastically reduce the amount of work as compared with the case where the abnormality diagnosis of the monitoring target is individually performed as in the related art.

[Brief description of the drawings]

FIG. 1 is a system configuration diagram showing an embodiment of an abnormality diagnosis device according to the present invention.

FIGS. 2A and 2B are graphs showing an acoustic signal measured by the abnormality diagnostic apparatus according to the present invention, wherein FIG. 2A shows a time waveform of the acoustic signal, and FIG. 2B shows a state of the acoustic signal plotted on polar coordinates. ing.

FIG. 3 is an explanatory diagram showing an example of an SDP expression of an acoustic signal plotted line-symmetrically with respect to a plurality of symmetry axes of polar coordinates.

FIG. 4 is a table showing time waveforms and SDP expressions of sound signals of a normal FFU and an abnormal FFU measured by the abnormality diagnosis device and method according to the present invention.

FIG. 5 is a table showing time waveforms and SDP expressions of sound signals of a normal FFU and an abnormal FFU measured by the abnormality diagnosis device and method according to the present invention.

FIG. 6 is a table showing time waveforms and SDP expressions of sound signals of a normal FFU and an abnormal FFU measured by the abnormality diagnosis apparatus and method according to the present invention.

FIG. 7 is an explanatory diagram showing an embodiment in which the abnormality diagnosis device according to the present invention is applied to a large number of FFUs installed in a wide facility space on the ceiling of a clean room.

FIG. 8 is a flowchart showing an example of an abnormality diagnosis method according to the present invention.

FIG. 9 is a flowchart illustrating an example of an abnormality diagnosis method according to the present invention.

FIG. 10 is a flowchart illustrating an example of an abnormality diagnosis method according to the present invention.

FIG. 11 is an explanatory diagram showing a case where a pattern recognition technique is applied to the abnormality diagnosis apparatus and method according to the present invention.

[Explanation of symbols]

 12 FFU 14 Microphone 16 Sound level meter 18 A / D converter 19 Signal processing unit 20 Display device 20a Normal pattern 20b Abnormal pattern

──────────────────────────────────────────────────続 き Continued on the front page (58) Field surveyed (Int.Cl. ⁷ , DB name) G01H 17/00 G01M 19/00

Claims (7)

(57) [Claims] 1 or 2 installed in an arbitrary equipment space
The sound generated by the monitoring target is collected by a sound collection device
Then, based on the acoustic signal, the abnormality of the monitored object is diagnosed.
With an abnormal diagnosis deviceThere: is there The change in the sound signal sampled over a time interval
Standardizing means for setting a moving width as a reference value;Acoustic signal sampled at a certain time (t)
The relative magnitude (A) of (X (t)) with respect to the reference value
A certain time in a polar coordinate system having a radius of (i))
The acoustic signal (X
The relative magnitude of (t + τ)) to the reference value is declination
(Θ (i), Φ (i)) gives the polar coordinates
At least one line-symmetric about at least one start line of the system
And two dots (A (i), Θ (i)) (A (i), Φ
Plotting means for plotting (i)) A dot pattern plotted by the plotting means
Display means for visually displaying the dot pattern; and a dot pattern displayed by the display means.
Diagnosing an abnormality of the monitored object according to
, An abnormality diagnosis device. 2. One or two installed in an arbitrary equipment space
The sound generated by the monitoring target is collected by a sound collection device
Then, based on the acoustic signal, the abnormality of the monitored object is diagnosed.
Abnormal diagnostic deviceAnd: is there The change in the sound signal sampled over a time interval
Standardizing means for setting a moving width as a reference value;Acoustic signal sampled at a certain time (t)
The relative magnitude (A) of (X (t)) with respect to the reference value
A certain time in a polar coordinate system having a radius of (i))
The acoustic signal (X
The relative magnitude of (t + τ)) to the reference value is declination
(Θ (i), Φ (i)) gives the polar coordinates
At least one line-symmetric about at least one start line of the system
And two dots (A (i), Θ (i)) (A (i), Φ
Plotting means for plotting (i)) A dot pattern plotted by the plotting means
A pattern recognizing means for recognizing the pattern;
Dot pattern and the previously acquired normal dot pattern
The dot pattern at the time of diagnosis and the correctness.
There is a deviation of more than a predetermined value from the regular dot pattern
The monitoring object is diagnosed as abnormal.
And a comparative diagnostic means. 3. One or two installed in an arbitrary equipment space
A method of diagnosing an abnormality of the monitored object based on the sound signal by collecting the sound generated by the monitored object by a sound collecting device: an acoustic signal sampled over a certain time interval Setting the fluctuation range of the sound signal as a reference value; and an acoustic signal sampled at a certain time (t).
The relative magnitude (A) of (X (t)) with respect to the reference value
A certain time in a polar coordinate system having a radius of (i))
The acoustic signal (X
The relative magnitude of (t + τ)) to the reference value is declination
(Θ (i), Φ (i)) gives the polar coordinates
At least one line-symmetric about at least one start line of the system
And two dots (A (i), Θ (i)) (A (i), Φ
(I)) plotting ; and repeating the step of plotting dots in the polar coordinate system over a predetermined period of time, and diagnosing an abnormality of the monitoring target based on a dot pattern obtained as a result. A method for diagnosing an abnormality, comprising: 4. The step of diagnosing an abnormality of the monitored object includes visually displaying an obtained dot pattern, and comparing a previously acquired normal dot pattern with a diagnostic dot pattern. The abnormality diagnosis method according to claim 3, wherein the abnormality is diagnosed for the monitoring target. 5. The step of diagnosing an abnormality of the monitoring target includes pattern recognition of the obtained dot pattern, pattern comparison between a previously obtained normal dot pattern and a diagnostic dot pattern, and When there is a deviation equal to or more than a predetermined value between the dot pattern at the time of the diagnosis and the dot pattern at the normal time, the monitoring target is diagnosed as being abnormal. Item 3. The abnormality diagnosis method according to Item 3. 6. A method of collecting sounds generated by a plurality of monitoring objects installed in a large facility space by a sound collection device and identifying an abnormal operation monitoring object based on the sound signals. Dividing and narrowing down the global facility space into a plurality of smaller facility spaces each containing a plurality of monitored objects; and in each of the smaller facility spaces over a certain time interval.
Setting the fluctuation range of the sampled sound signal as a reference value; the sound signal sampled at a certain time (t)
The relative magnitude (A) of (X (t)) with respect to the reference value
A certain time in a polar coordinate system having a radius of (i))
The acoustic signal (X
The relative magnitude of (t + τ)) to the reference value is declination
(Θ (i), Φ (i)) gives the polar coordinates
At least one line-symmetric about at least one start line of the system
And two dots (A (i), Θ (i)) (A (i), Φ
(I)) plotting ; and repeating the step of plotting dots in the polar coordinate system over a predetermined period of time. Based on the resulting dot pattern, a small area including an abnormal operation monitoring target is included. Specifying a facility space; and specifying a monitoring target of an abnormal operation from a plurality of monitoring targets included in the small-sized facility space specified in the step. Abnormal diagnosis method. 7. Before specifying an object to be monitored for abnormal operation,
The necessary steps for dividing the small facility space containing the monitoring target of abnormal operation into smaller facility spaces again and narrowing down the facility space containing the monitoring target of abnormal operation are performed as many times as necessary. 7. The abnormality diagnosis method according to claim 6, wherein the method is repeated.