JP7814173B2 - Rotation angle sensor abnormality diagnosis device - Google Patents

Description

The present invention relates to a device for diagnosing abnormalities in rotation angle sensors.

In recent years, with the electrification of automobiles, vehicles using permanent magnet synchronous motors (PMSMs) for their main and auxiliary machinery (brakes, power steering, etc.) have become widespread. Information from rotation angle sensors is essential for PMSM control. To improve the reliability of rotation angle sensors, it is necessary to ensure the validity of the rotation angle sensor signal when using their values.

Patent Document 1 describes a technology that detects the phase shift with respect to an excitation signal, which is the input signal of a resolver (rotation angle sensor), based on the values of the excitation component, sine wave component, and cosine wave component sampled at multiple points in time during one cycle of the excitation signal. The phase shift is detected by detecting the presence or absence of a cosine wave component at a specified angle in the excitation signal.

Japanese Patent Application Laid-Open No. 2020-190533

An excitation signal is input to the resolver, but the technology described in Patent Document 1 assumes that there is no abnormality in the period of the resolver's excitation signal.

However, there is a need to further improve the reliability of the rotation angle sensor by detecting any abnormalities in the period of the resolver excitation signal and ensuring the validity of the rotation sensor signal.

The present invention was made in consideration of the above, and its purpose is to realize an abnormality diagnosis device for a rotation angle sensor that improves the accuracy of detecting abnormalities in the rotation angle sensor.

To achieve the above objectives, the present invention is configured as follows:

An abnormality diagnosis device for a rotation angle sensor that determines whether or not there is an abnormality in a rotation angle sensor provided in a motor based on a sine signal and a cosine signal output from the rotation angle sensor. The device includes: a data acquisition unit that acquires the sine signal and the cosine signal output from the rotation angle sensor; a square sum calculation unit that squares and adds the data of the sine signal and the cosine signal obtained by the data acquisition unit and calculates a square sum calculation result; and an abnormality determination unit that determines whether or not there is an abnormality in the rotation angle sensor based on the calculation result of the square sum calculation unit. The abnormality determination unit determines whether or not the periods of the sine signal and the cosine signal are abnormal based on the obtained peak period or period of zero value based on the square sum calculation result, and determines whether or not the amplitude values of the sine signal and the cosine signal are abnormal based on the obtained peak value based on the square sum calculation result.

According to the present invention, it is possible to realize an abnormality diagnosis device for a rotation angle sensor with improved accuracy in detecting abnormalities in the rotation angle sensor.
Problems, configurations, and effects other than those described above will become apparent from the following description of the embodiments.

1 is a schematic configuration diagram of a motor control device to which a resolver, which is a rotation angle sensor, is connected by an abnormality diagnosis device according to an embodiment of the present invention; 1 is a diagram showing a schematic configuration of an abnormality diagnosis device 400 according to an embodiment of the present invention. 10A and 10B are explanatory diagrams of a method for storing buffer data saved in a data buffer unit. 1 is a diagram showing the relationship between an excitation signal and amplitude-modulated sine and cosine components to explain the principles of the present invention; FIG. 10 is a graph for explaining the square sum characteristics of a sine component and a cosine component; 10 is a graph illustrating the relationship between an excitation signal and a square sum signal. 10 is a graph illustrating the relationship between an excitation signal and a square sum signal. 4 is a flowchart illustrating the operation of an abnormality determination unit in the abnormality diagnosis device.

Embodiments of the present invention will be described below with reference to the drawings. Please note that the embodiments of the present invention are not limited to the examples described below, and various modifications are possible within the scope of the technical concept.

Figure 1 is a schematic diagram of a motor control device 100 to which a resolver 320, which is a rotation angle sensor, is connected by an abnormality diagnosis device 400 according to one embodiment of the present invention.

In FIG. 1, motor control device 100 is connected to motor 300. Motor drive device 100 has a current control unit 110, an inverter 130, a rotational position detection unit 150, and an excitation unit 160. Battery 200 is the DC voltage source for motor drive device 100. DC power stored in battery 200 is converted into variable-voltage, variable-frequency three-phase AC power by inverter 130 of motor drive device 100. Inverter 130 supplies this three-phase AC power to motor 300.

Motor 300 is a synchronous motor that is driven to rotate by a supply of three-phase AC power. A resolver 320 is attached to motor 300 as a rotation angle sensor for rotor 310 to control the phase of the three-phase AC voltage to match the phase of the induced voltage of motor 300.

The resolver 320 receives the excitation signal (sinωt) from the excitation unit 160 and outputs amplitude-modulated two-phase signals Sn1 (sine wave signal (sinθ sinωt)) and Cn1 (cosine wave signal (cosθ sinωt)) that are 90° out of phase with each other (θ: motor rotation angle).

The current control unit 110 generates a drive signal using pulse width modulation (PWM) to follow the current command value generated by a higher-level control device (not shown). The current control unit 110 calculates three-phase motor voltage command values based on the motor current values (Iu, Iv, Iw) detected by the current sensor 170 and the rotation angle θ of the motor 300 detected by the rotational position detection unit 150.

The drive signal described above controls the on/off operation of the semiconductor switch elements that make up the main circuit of the inverter 130, causing the inverter 130 to output three-phase voltages (Vu, Vv, Vw) in response to the motor voltage command.

The signal input to the rotational position detection unit 150 in response to the two output signals (Sn1, Cn1) of the resolver 320 is sampled at timings that are odd multiples of π/2 (π/2, 3π/2, 5π/2, ...) in electrical angle terms in the excitation signal (sinωt), thereby removing modulation.

The rotational position detection unit 150 calculates the rotation angle θ of the motor 300 based on the two-phase signal from which the modulation has been removed. The abnormality diagnosis device 400 receives the output signals of the resolver 320 and detects the presence or absence of an abnormality in the resolver 320 based on these signals.

When the abnormality diagnosis device 400 detects an abnormality, it outputs an abnormality determination signal J1 to a higher-level control device (not shown) to notify the higher-level control device of an abnormality in the resolver 320.

Figure 2 shows the general configuration of an abnormality diagnosis device 400, one embodiment of the present invention.

In Figure 2, two amplitude-modulated output signals Sn1 and Cn1 from resolver 320 are input into abnormality diagnosis device 400. The two output signals Sn1 and Cn1 are first acquired by AD conversion unit (high-speed data acquisition unit) 401 at a sampling period six or more times the period of the excitation signal.

A sampling rate of 6 times or more is a value to ensure the minimum number of samples required to measure the peak period in the square sum calculation described below, but to improve the accuracy of waveform reproduction, it may be preferable to increase the sampling rate to 10 times or more.

The data Sn2 and Cn2 sampled by the AD conversion unit 401 is stored for one excitation period in the data buffer unit 402. At this time, the time of AD conversion, for example, the free-running timer value T1, is also stored in the buffer unit 402 along with each signal.

As will be explained later, the buffer data stored in the data buffer unit 402 can be managed by index by storing each data as an array.

The two output signals Sn2B and Cn2B from the data buffer unit 402 are subjected to a square sum calculation for each index by the square sum calculator 403, and the data is similarly stored as an array.

The peak index detection unit 404 extracts the amplitude peak from the data SQ1 obtained by the square sum calculation performed by the square sum calculation unit 403. The free-running timer value T1 stored in the data buffer unit 402 is output to the peak index detection unit 404 as T1B.

As an example of a method for detecting amplitude peaks, the following peak value search method may be used.

That is, in the peak value search method, for example, three pieces of data are extracted using consecutive indexes, and when the data are designated as data 1, data 2, and data 3, data 2 may be selected if the following condition (1) is met: Condition (1): data 1<data 2 & data 2>data 3
Using the index of the data obtained under the above condition (1), amplitude peak values Pk1 and Pk2, which are consecutive peak values for calculating the period, are extracted. Also, AD conversion time data Tpk1 and Tpk2 at which the amplitude peak values occurred are extracted.

The amplitude peak value Pk1, amplitude peak value Pk2, AD conversion time data Tpk1, and AD conversion time data Tpk2 are output to the abnormality determination unit 405. The abnormality determination unit 405 then derives the period of the square sum data SQ1 (square sum period fsqr) using the following equation (1):

Square sum period fsqr=1/(Tpk2-Tpk1) (1)
The abnormality determination unit 405 compares the square sum period fsqr with the excitation period expected by the system, determines whether or not an abnormality exists, and outputs an abnormality determination result J1. The abnormality determination result J1 can be notified by an appropriate display device or the like.

Figure 3 is an explanatory diagram of how buffer data is stored in the data buffer unit 402.

A predetermined amount of data Sn2 sampled by the AD conversion unit 401 is stored in chronological order in the sine component table 501. An index is assigned to each piece of data to facilitate data search. When the buffer data is filled with a predetermined amount of data, the oldest data is overwritten and saved in chronological order.

Similarly, a predetermined amount of data Cn2 sampled by the AD conversion unit 401 is stored in chronological order in the cosine component table 502.

Furthermore, the timestamps sampled by the AD conversion unit 401 are stored in chronological order for a specified amount of data in the AD conversion timestamp table 503.

In addition, the square sum calculation unit 403 stores a predetermined amount of square sum data SQ1 of the sine and cosine components for each index in a time series in the square sum table 504.

In this embodiment, the data size stored in the data buffer unit 402 is set to 10, but in actual operation, the size can be changed depending on the number of AD conversion samples in one excitation period.

Figure 4 is a diagram illustrating the relationship between the excitation signal (sinωt) and the amplitude-modulated sine component Sn1 and cosine component Cn1 to explain the principles of the present invention. Figure 4 shows an enlarged view of the amplitude-modulated sine component Sn1 and cosine component Cn1 over several periods of the excitation signal when the excitation signal (sinωt) is applied to the resolver 320, at an arbitrary time period.

In Figure 4, the excitation signal frequency and the frequencies of the amplitude-modulated sine component Sn1 and cosine component Cn1 are basically the same frequency due to the circuit configuration (although there may be a slight phase difference).

Therefore, by observing the frequencies of the sine component Sn1 and the cosine component Cn1, the integrity of the excitation cycle should be observable.

However, as can be seen from the enlarged waveform in Figure 4, there are sections where the amplitude is small depending on the observation timing (the amplitude of the sine component, Sn1, is small), making it difficult or impossible to observe the frequency of the waveform in these sections.

The example shown in Figure 4 shows observations at times when the amplitude of the sine component, Sn1, decreases, but there are also times when the amplitude of the cosine component Cn1 decreases as well.

Next, we will explain the square-sum characteristics of the sine component Sn1 and cosine component Cn1 of the two amplitude-modulated waves output by resolver 320, which is the basis of this invention.

Figure 5 is a graph illustrating the square sum characteristics of the sine component Sn1 and the cosine component Cn1.

The excitation signal (sinωt) input to the resolver 320 is set to a predetermined amplitude and frequency.

On the other hand, the sine component Sn1 and cosine component Cn1 of the signal output from resolver 320 after being amplitude modulated by the excitation signal (sinωt) have waveforms whose amplitudes vary depending on the angular position of resolver 320.

The signal components are sine component Sn1 = sinθsinωt and cosine component Cn1 = cosθsinωt.

In particular, the sine component Sn1 has an amplitude of approximately 0 when the angle of the resolver 320 is 0 degrees, 180 degrees, etc. Conversely, the cosine component Cn1 has an amplitude of approximately 0 when the angle of the resolver 320 is 90 degrees, 270 degrees, etc. This corresponds to the amplitude of the sine and cosine components of the angle.

On the other hand, the signal Sqr1 (the square sum calculation result) calculated by taking the square sum of the sine component Sn1 and the cosine component Cn1 has the components shown in the following equation (2).

Sqr1=(sinωt) ² (sinθ ² +cosθ ² ) ...(2)
The following formula (3) is established from the basic formula of trigonometric functions.

Sqr1=(sinωt) ² ...(3)
From the above equation (3), unlike the sine component Sn1 and the cosine component Cn1, the sum-of-squares signal Sqr1 has a sine wave characteristic of constant amplitude regardless of the angle of the resolver 320.

Furthermore, as shown in Figures 6 and 7, the squared sum signal Sqr1 is a sine wave of constant amplitude that is synchronized (linked) with the excitation signal (sinωt), and because it is a double angle equation, the frequency fsqr of the signal Sqr1 is exactly twice the excitation signal frequency fexc.

The present invention uses a signal Sqr1, which is a square sum of a constant amplitude and has twice the period of the excitation signal, to diagnose abnormalities in the resolver 320, which is a rotation angle sensor.

Figure 8 is an operational flowchart of the abnormality determination unit 405 in the abnormality diagnosis device 400.

In step S1 of Figure 8, the abnormality determination unit 405 begins operation to detect whether or not there is an abnormality in the resolver 320.

Next, in step S2, the abnormality determination unit 405 acquires the amplitude peak value Pk1 and amplitude peak value Pk2 acquired by the peak index detection unit 404.

Next, in step S3, the abnormality determination unit 405 acquires the AD conversion times tpk1 and tpk2 (corresponding to the AD conversion time data Tpk1 and Ttpk2) corresponding to the indexes of the amplitude peak values Pk1 and Pk2 acquired by the peak index detection unit 404.

Next, in step S4, the abnormality determination unit 405 calculates the square sum signal period fsqr from the AD conversion times tpk1 and tpk2 acquired in S3 (the calculation formula is described above).

Next, in step S5, the abnormality determination unit 405 determines whether half the value of the square sum signal period fsqr obtained in step S4 is less than the upper limit value fexcmax of the excitation signal period input to the resolver 320 from the excitation unit 160, or exceeds the lower limit value fexcmin, to determine whether there is an abnormality in the excitation signal period.

If the condition is met in step S5, it is determined in step S6 that there is no abnormality in the excitation signal period, and the period abnormality flag is set to 0 (no abnormality).

On the other hand, if the condition in step S5 is not met, in step S7 it is determined that there is an abnormality in the excitation signal period, and the period abnormality flag is set to 1 (abnormal).

Furthermore, in step S8, the abnormality determination unit 405 determines whether the amplitude peak value Pk1 or amplitude peak value Pk2 of the square sum signal acquired in step S2 is less than the upper limit value Pkmax or greater than the lower limit value Pkmin, and determines whether there is an abnormality in the amplitude values of the output signals Sn1 and Cn1 of the resolver 320.

If the above condition is met in step S8, the amplitude abnormality flag is set to 0 (no abnormality) in step S9, and the series of operations ends.

On the other hand, if the above condition is not met in step S8, the amplitude abnormality flag is set to 1 (abnormality present) in step S10, and the series of operations is stopped.

Note that steps S5 to S7 and steps S8 to S10 may be completed independently.

In addition, the abnormality judgment J1 of the abnormality judgment unit 405 can be displayed on an appropriate display device (not shown).

As described above, according to the present invention, the output signals Sn1 and Cn1 of the resolver 320, which is a rotation angle sensor, are summed by squares to obtain sum-of-squares data, and the peak values Pk1 and Pk2 of the sum-of-squares data are detected to determine whether the obtained peak period is abnormal, and whether peak value Pk1 or peak value Pk2 is abnormal is also determined.

This makes it possible to realize a rotation angle sensor abnormality diagnosis device with improved accuracy in detecting abnormalities in the rotation angle sensor.

Furthermore, by taking the square sum of the output signals Sn1 and Cn1 of the resolver 320 to obtain a square sum signal, sine waves and cosine waves of a predetermined amplitude can be obtained regardless of the angular position. This makes it possible to determine whether there are any abnormalities in the excitation signal, sine wave signal, and cosine signal regardless of the angular position.

In the above-described embodiment, the period of the excitation signal sinωt is calculated from the peak values Pk1 and Pk2 of the square sum signal, but it is also possible to detect the zero value of the square sum signal and calculate the period of the zero value as the period of the excitation signal sinωt.

100: Motor drive device, 110: Current control unit, 130: Inverter, 150: Rotational position detection unit, 160: Excitation unit, 200: Battery, 300: Motor, 310: Rotor, 320: Resolver, 400: Abnormality diagnosis device, 401: AD conversion unit, 402: Data buffer unit, 403: Square sum calculation unit, 404: Peak index detection unit, 405: Abnormality determination unit, 501: Sine component table, 502: Cosine component table, 503: AD conversion timestamp table, 504: Square sum table

Claims (1)

1. A rotation angle sensor abnormality diagnosis device for determining whether or not an abnormality exists in a rotation angle sensor provided in a motor based on a sine signal and a cosine signal output from the rotation angle sensor,
a data acquisition unit that acquires the sine signal and the cosine signal output from the rotation angle sensor;
a square sum calculation unit that squares and adds the sine signal data and the cosine signal data obtained by the data acquisition unit, and calculates a square sum calculation result;
an abnormality determination unit that determines whether or not there is an abnormality in the rotation angle sensor based on the calculation result of the square sum calculation unit;
Equipped with
The abnormality determination unit
determining whether the periods of the sine signal and the cosine signal are abnormal based on the peak period or the period of the zero value obtained based on the result of the square sum calculation;
The rotation angle sensor abnormality diagnostic device is characterized in that it determines whether the amplitude values of the sine signal and the cosine signal are abnormal or not based on the peak value obtained from the result of the square sum calculation.