PWM control device and PWM control method

Abstract

To make it possible to detect abnormality of PWM CONTROL generated inside a PWM control device, while being able to detect abnormality even when a duty command is 0% or 100%.SOLUTION: A PWM control device 100 comprises output pulse width acquisition means 12 for acquiring an output pulse width Wact of a PWM control signal SW, command pulse width calculation means 13 for calculating a command pulse width Wref of a duty command Dref, and abnormality determination means 17 for determining presence / absence of abnormality of PWM control by comparing the output pulse width Wact with the command pulse width Wref. The output pulse width acquisition means 12 measures the output pulse width of the PWM control signal SW, and corrects a measured value of the output pulse width to make the corrected measured value of the output pulse width be the output pulse width Wact, when it is determined that there is no output pulse.SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] This application relates to a PWM control device and a PWM control method.

Background Art

[0002] As a control method for a load device such as a flow control valve or a motor, PWM (Pulse Width Modulation) control is widely used. In PWM control, a PWM control signal generated by a PWM control device is transmitted to a drive circuit to generate a drive signal corresponding to the PWM control signal. The generated drive signal is transmitted to the load device, and the load device is driven by the drive signal. In such PWM control, it is necessary to check whether there is an abnormality in the PWM control being executed. Here, as an abnormality detection device for a conventional control circuit, a check pulse with a fixed period is input to an output transistor that controls the operation of the load, and an inverted pulse of this check pulse is obtained, and the pulse width of the check pulse and the pulse width of the inverted pulse are compared to detect an abnormality (see, for example, Patent Document 1). Also, there is a method of detecting an abnormality in PWM control by sequentially comparing the level of the PWM control signal output from the PWM control device and the level of the drive signal output from the drive circuit (see, for example, Patent Document 2).

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Patent Document 2

Summary of the Invention

Problems to be Solved by the Invention

[0004] However, when applying a method similar to that of an anomaly detection device such as Patent Document 1 to PWM control, when the output pulse is seemingly nonexistent, such as when the duty cycle command is 0% or 100%, it is not possible to obtain the pulse width for comparison and therefore not be able to detect anomalies. Furthermore, while methods that compare the level of the PWM control signal and the level of the drive signal, as disclosed in Patent Document 2, can detect abnormalities even when the duty cycle command is 0% or 100%, they cannot detect abnormalities within the PWM control device because they use the PWM control signal output from the PWM control device as the criterion for abnormality detection. Therefore, they cannot determine whether the PWM control signal is output from the PWM control device in accordance with the duty cycle command defined in the PWM control device. This application discloses a technology for solving the above-mentioned problems, and aims to provide a PWM control device and a PWM control method that can detect abnormalities even when the duty cycle is 0% or 100%, and can also detect abnormalities in PWM control that occur within the PWM control device. [Means for solving the problem]

[0005] The PWM control device disclosed in this application includes a control command generation means for generating a PWM control command, a control signal generation means for generating a PWM control signal for controlling a controlled device based on the PWM control command, an output pulse width acquisition means for acquiring the output pulse width of the PWM control signal, a command pulse width calculation means for calculating the command pulse width which is the pulse width of the PWM control command, and an abnormality determination means for determining whether or not there is an abnormality in the PWM control by comparing the output pulse width and the command pulse width. The output pulse width acquisition means measures an output pulse width measurement value which is the measured value of the width of the output pulse of the PWM control signal based on the edge of the PWM control signal, and determines whether or not there is an output pulse of the PWM control signal. If it is determined that there is no output pulse, it corrects the output pulse width measurement value and uses the corrected output pulse width measurement value as the output pulse width.

[0006] Furthermore, another PWM control device disclosed in this application is a PWM control device that outputs a PWM control signal to a controlled device having an electric motor that operates by being supplied with power according to a PWM control signal, and comprises control command generation means for generating a PWM control command, control signal generation means for generating a PWM control signal based on the PWM control command, at least one of output pulse width acquisition means for acquiring the output pulse width of the PWM control signal and command pulse width calculation means for calculating the command pulse width which is the pulse width of the PWM control command, abnormality determination means for determining whether or not there is an abnormality in the PWM control, and rotation state acquisition means for acquiring the rotation state of the electric motor, wherein the abnormality determination means has a pulse width fixing time measurement unit that measures the pulse width fixing time which is the time during which the output pulse width or command pulse width remains constant, and determines whether or not there is an abnormality in the PWM control based on a comparison of the pulse width fixing time with a time determined based on the rotation state and whether or not there is a time variation of the ideal value of the PWM control command.

[0007] Furthermore, the PWM control method disclosed in this application comprises a control command generation step of generating a PWM control command, a control signal generation step of generating a PWM control signal that controls a controlled device based on the PWM control command, an output pulse width acquisition step of acquiring the output pulse width of the PWM control signal, a command pulse width calculation step of calculating the command pulse width which is the pulse width of the PWM control command, and an abnormality determination step of determining whether or not there is an abnormality in the PWM control by comparing the output pulse width and the command pulse width. In the output pulse width acquisition step, the output pulse width measurement value, which is the measured value of the width of the output pulse of the PWM control signal, is measured based on the edge of the PWM control signal, and the presence or absence of an output pulse of the PWM control signal is determined. If it is determined that there is no output pulse, the output pulse width measurement value is corrected, and the corrected output pulse width measurement value is taken as the output pulse width.

[0008] Furthermore, another PWM control method disclosed in this application is a PWM control method that outputs a PWM control signal to a control device having an electric motor that operates by being supplied with power according to a PWM control signal, and comprises a control command generation step of generating a PWM control command, a control signal generation step of generating a PWM control signal based on the PWM control command, at least one of an output pulse width acquisition step of acquiring the output pulse width of the PWM control signal and a command pulse width calculation step of calculating a command pulse width which is the pulse width of the PWM control command, an abnormality determination step of determining whether or not there is an abnormality in the PWM control, and a rotation state acquisition step of acquiring the rotation state of the electric motor, wherein in the abnormality determination step, the pulse width fixing time which is the time during which the output pulse width or command pulse width remains constant is measured, and the presence or absence of an abnormality in the PWM control is determined based on a comparison of the pulse width fixing time with a time determined based on the rotation state, and the presence or absence of time fluctuation of the ideal value of the PWM control command. [Effects of the Invention]

[0009] According to the PWM control device or PWM control method disclosed herein, abnormalities can be detected even when the duty cycle command is 0% or 100%, and abnormalities in PWM control that occur within the PWM control device can be detected. [Brief explanation of the drawing]

[0010] [Figure 1] This is a block diagram showing the configuration of the PWM control device in Embodiment 1. [Figure 2] This is a block diagram showing the configuration of the control command generation means according to Embodiment 1. [Figure 3] This is a block diagram showing the configuration of the abnormality determination means according to Embodiment 1. [Figure 4] This is a block diagram showing the configuration of the output pulse width acquisition means according to Embodiment 1. [Figure 5] This figure shows the relationship between the PWM control signal and the output pulse width measurement value according to Embodiment 1, and is a figure showing the case where no correction is made to the output pulse width measurement value. [Figure 6A]It is a diagram showing the relationship between the PWM control signal and the output pulse width measurement value according to Embodiment 1, and is a diagram showing the case of correcting the output pulse width measurement value. [Figure 6B] It is a diagram for explaining the determination time of the output pulse absence determination in Embodiment 1. [Figure 7] It is a diagram showing an example of the hardware configuration for realizing the control unit according to Embodiment 1. [Figure 8] It is a flowchart showing the operation of the PWM control device in Embodiment 1. [Figure 9] It is a flowchart showing the output pulse width acquisition step according to Embodiment 1. [Figure 10] It is a flowchart showing the abnormality determination step according to Embodiment 1. [Figure 11] It is a block diagram showing the configuration of the power conversion device according to Embodiment 2. [Figure 12] It is a block diagram showing the configuration of the control unit according to Embodiment 2. [Figure 13] It is a block diagram showing the configuration of the control command generation means according to Embodiment 2. [Figure 14] It is a block diagram showing the configuration of the rotation state acquisition means according to Embodiment 2. [Figure 15] It is a block diagram showing the configuration of the abnormality determination means according to Embodiment 2. [Figure 16] In Embodiment 2, it is a diagram for comparing the time waveform of the electrical angle and the time waveform of the PWM control signal on the upper side of the U phase. [Figure 17] It is a flowchart showing the operation of the PWM control device in Embodiment 2. [Figure 18] It is a flowchart showing the abnormality determination step according to Embodiment 2. [Figure 19] It is a block diagram showing the configuration of the control unit according to Embodiment 3. [Figure 20] It is a block diagram showing the configuration of the abnormality determination means according to Embodiment 3. [Figure 21] It is a flowchart showing the operation of the PWM control device in Embodiment 3. [Figure 22] It is a flowchart showing the abnormality determination process according to Embodiment 3. [Figure 23] It is a block diagram showing the configuration of the abnormality determination means according to Embodiment 4. [Figure 24] It is a flowchart showing the abnormality determination process according to Embodiment 4.

Mode for Carrying Out the Invention

[0011] Embodiment 1. Hereinafter, the configuration and operation of the PWM control device according to Embodiment 1 of Embodiment 1 will be described based on FIGS. 1 to 10.

[0012] <Configuration of PWM Control Device> FIG. 1 is a block diagram showing the configuration of the PWM control device in Embodiment 1. The PWM control device 100 includes a control unit 10, and transmits the PWM control signal SW generated by the control unit 10 to the controlled device 900. The controlled device 900 is controlled according to the PWM control signal. The controlled device 900 includes a drive circuit that generates a drive signal according to the PWM control signal, and a load device such as an electric motor that operates by the drive signal. However, the internal configuration of the controlled device 900 is omitted in FIG. 1.

[0013] <Functional Blocks of Control Unit> The control unit 10 includes a control command generation means 11 that generates a duty command Dref as a PWM control command and outputs the duty command Dref to a command pulse width calculation means 13 and a control signal generation means 19; a control signal generation means 19 that generates a PWM control signal SW based on the duty command Dref and outputs the PWM control signal SW to a controlled device 900 and an output pulse width acquisition means 12; an output pulse width acquisition means 12 that acquires the output pulse width Wact, which is the pulse width of the PWM control signal SW, and outputs the output pulse width Wact to an abnormality determination means 17; a command pulse width calculation means 13 that calculates the command pulse width Wref based on the duty command Dref and outputs the command pulse width Wref to an abnormality determination means 17; and an abnormality determination means 17 that determines whether there is an abnormality from the comparison result between the output pulse width Wact and the command pulse width Wref and outputs the result as a state determination result ErrSt. The control unit 10 also includes an output stop instruction means 18 that receives a state determination result ErrSt from the abnormality determination means 17 and outputs a stop instruction signal Stp to the control signal generation means 19 according to the state determination result ErrSt.

[0014] The control command generation means 11 will now be described. Figure 2 is a block diagram showing the configuration of the control command generation means according to Embodiment 1. The control command generation means 11 has a duty cycle command calculation unit 11a. The duty cycle command calculation unit 11a calculates a duty cycle command Dref for generating a PWM control signal SW based on a control command from a higher-level controller or system (not shown), input from a user, or feedback from the controlled device 900, and outputs the calculated duty cycle command Dref.

[0015] Details of the output pulse width acquisition means 12 will be described later.

[0016] The command pulse width calculation means 13 calculates the desired pulse width of the PWM control signal SW. This "desired pulse width" is the command pulse width Wref, and the command pulse width calculation means 13 calculates the command pulse width Wref based on the duty cycle command Dref.

[0017] The abnormality determination means 17 will now be described. Figure 3 is a block diagram showing the configuration of the abnormality determination means according to Embodiment 1. The abnormality determination means 17 includes a pulse width comparison unit 17a and a difference abnormality determination unit 17b. The pulse width comparison unit 17a receives the command pulse width Wref and the output pulse width Wact as input. The pulse width comparison unit 17a compares the command pulse width Wref and the output pulse width Wact, calculates the difference WDiff between the two, and outputs the difference WDiff to the difference abnormality determination unit 17b. The difference abnormality determination unit 17b determines whether the difference WDiff is greater than or equal to the threshold Dth and outputs a state determination result ErrSt based on the result. If the difference WDiff is greater than or equal to the threshold Dth, the difference abnormality determination unit 17b determines that there is an abnormality in the PWM control and sets the state determination result ErrSt to "abnormal". If the difference WDiff is less than the threshold Dth, it determines that the PWM control is operating normally and sets the state determination result ErrSt to "normal".

[0018] The output pulse width Wact is the pulse width of the PWM control signal SW output by the control signal generation means 19. The control signal generation means 19 also transmits the generated PWM control signal SW to the output pulse width acquisition means 12 and the controlled device 900, respectively. Therefore, the PWM control signal SW received by the output pulse width acquisition means 12 does not pass through the inside of the controlled device 900, nor does it pass through the transmission path between the PWM control device 100 and the controlled device 900. For this reason, if there is a difference between the output pulse width Wact acquired by the output pulse width acquisition means 12 and the command pulse width Wref, the cause of this difference is that it is occurring within the PWM control device 100. For this reason, if the difference WDiff is large (greater than or equal to the threshold Dth), it indicates that an abnormality has occurred within the PWM control device 100.

[0019] When determining the presence or absence of an anomaly by comparing two "pulse widths," as in the anomaly determination means 17 of Embodiment 1, there is no need to provide additional dedicated circuitry. However, when comparing two "signal levels," as in the technology described in Patent Document 2, additional hardware such as dedicated circuitry is required. While it is possible to compare signal levels using software without adding hardware, such a configuration has the problem of increasing the processing load on the computing device. The anomaly determination means 17 of Embodiment 1 does not increase the processing load on the computing device and does not require the addition of dedicated circuitry or the like.

[0020] The output stop instruction means 18 stops the generation of the PWM control signal SW by the control signal generation means 19 according to the state of PWM control indicated by the state determination result ErrSt. If the state determination result ErrSt is "normal", the output stop instruction means 18 does nothing, and if the state determination result ErrSt is "abnormal", it outputs a stop instruction signal Stp to the control signal generation means 19. Upon receiving the stop instruction signal Stp, the control signal generation means 19 stops the generation and output of the PWM control signal SW.

[0021] The control signal generation means 19 generates and outputs a PWM control signal SW for controlling the controlled device 900 in accordance with the duty cycle command Dref. The control signal generation means 19 also generates and outputs a PWM control signal SW based on, for example, the PWM pulse period. The generation of the PWM control signal SW by the control signal generation means 19 is performed by comparing the duty cycle command Dref with the carrier wave. More specifically, the PWM control signal SW is generated using, for example, a triangular wave comparison method using a triangular wave having the shape of an isosceles triangle with equal rising and falling speeds as the carrier, or a sawtooth wave comparison method using a sawtooth wave that periodically repeats linear increases and sharp drops as the carrier. As described above, when the control signal generation means 19 receives a stop instruction signal Stp from the output stop instruction means 18, it stops generating and outputting the PWM control signal SW. The PWM pulse period may be predetermined or set according to the state of the load device.

[0022] The output pulse width acquisition means 12 will now be described. Figure 4 is a block diagram showing the configuration of the output pulse width acquisition means according to Embodiment 1. The output pulse width acquisition means 12 includes an output pulse edge detection unit 12a that notifies the detection of an edge of the output pulse of the PWM control signal SW, an output pulse width measurement unit 12b that measures the output pulse width of the PWM control signal SW and outputs it to the output pulse width correction unit 12f as the output pulse width measurement value Wmon, an edge time acquisition unit 12c that acquires the time of the edge timing using a time counter and outputs it to the pulse presence / absence determination unit 12d as the edge time EdgT, a pulse presence / absence determination unit 12d that determines the presence or absence of an output pulse in the PWM pulse period and outputs the result to the output pulse width correction unit 12f as the pulse presence / absence determination result EdgR, a signal level acquisition unit 12e that acquires the signal level at the edge timing of the PWM control signal SW and outputs it to the output pulse width correction unit 12f as the signal level EdgLev, and an output pulse width correction unit 12f that corrects the output pulse width measurement value Wmon according to the pulse presence / absence determination result EdgR, or outputs it as the output pulse width Wact without correction.

[0023] The output pulse edge detection unit 12a receives the PWM control signal SW and detects the rising and falling edges of the output pulses of the PWM control signal SW. The output pulse edge detection unit 12a outputs an edge detection signal Edg at the timing (edge ​​timing) when it detects an edge of the PWM control signal SW as described above. The edge detection signal Edg is input to the output pulse width measurement unit 12b, the edge time acquisition unit 12c, and the signal level acquisition unit 12e. In this way, the output pulse edge detection unit 12a notifies of the detection of an edge of the output pulse of the PWM control signal SW by outputting the edge detection signal Edg. The edge detection signal Edg is output at edge timing, and the receiving side of the edge detection signal Edg receives the edge detection signal Edg at edge timing.

[0024] The output pulse width measurement unit 12b receives the edge detection signal Edg and operates a counter at the edge timing of the PWM control signal SW to measure the output pulse width of the PWM control signal SW from the count value from rising to falling edge. The output pulse width measurement unit 12b outputs the measured output pulse width as the output pulse width measurement value Wmon. In Embodiment 1, the output pulse width of the PWM control signal SW is measured from the count value from rising edge to falling edge, but it may also be configured to measure the output pulse width of the PWM control signal SW from the count value from falling edge to rising edge.

[0025] Furthermore, if the control unit 10 is configured as a microcomputer, for example, the output pulse edge detection unit 12a and the output pulse width measurement unit 12b can be implemented by utilizing the input timer function of the microcomputer.

[0026] The edge time acquisition unit 12c receives the edge detection signal Edg and operates a time counter at the edge timing to acquire the edge timing. The edge time acquisition unit 12c outputs the acquired time as the edge time EdgT of the output pulse of the PWM control signal SW.

[0027] The pulse presence / absence determination unit 12d determines the presence or absence of an output pulse of the PWM control signal SW based on the elapsed time since the last edge timing. More specifically, the pulse presence / absence determination unit 12d receives the edge time EdgT and measures the elapsed time since the last edge timing, i.e., the last edge detection time, from the difference between the current time and the edge time EdgT. The pulse presence / absence determination unit 12d determines that there is no output pulse if the elapsed time since the last edge timing is equal to or greater than the PWM pulse period. The pulse presence / absence determination unit 12d performs the above determination at the PWM pulse period. That is, the pulse presence / absence determination unit 12d determines the presence or absence of an output pulse for each PWM pulse period. The pulse presence / absence determination unit 12d outputs the result of the determination as the pulse presence / absence determination result EdgR.

[0028] The pulse presence / absence determination unit 12d may be operated at a different period and timing than the PWM pulse period. In this case, the determination period for the presence or absence of an output pulse (the period subject to determination) should be set to a value corresponding to the execution period and execution timing of the pulse presence / absence determination unit 12d. This allows the execution of the determination by the pulse presence / absence determination unit 12d to be set without being limited to the PWM pulse period.

[0029] The signal level acquisition unit 12e receives the edge detection signal Edg, acquires the level of the PWM control signal SW at the edge timing as "high" or "low", and outputs it as the signal level EdgLev at the time of edge detection.

[0030] The output pulse width correction unit 12f receives the output pulse width measurement value Wmon, the pulse presence / absence determination result EdgR, and the signal level EdgLev at the time of edge detection as input. If the output pulse width correction unit 12f determines that there is an output pulse (the pulse presence / absence determination result EdgR is "pulse present"), it outputs the output pulse width measurement value Wmon as is without correction. If the output pulse is determined to be absent (the pulse presence / absence determination result EdgR is "no pulse"), it corrects the output pulse width measurement value Wmon and outputs it as the output pulse width Wact.

[0031] The method for correcting the output pulse width by the output pulse width correction unit 12f is as follows: When the signal level EdgLev is "low", the output pulse width correction unit 12f corrects the output pulse width measurement value Wmon to zero and outputs the output pulse width Wact as "output pulse width Wact value = 0". When the signal level EdgLev is "high", the output pulse width correction unit 12f corrects the output pulse width measurement value Wmon to the same value as the PWM pulse period and outputs the output pulse width Wact as "output pulse width Wact value = PWM pulse period".

[0032] In Embodiment 1, the signal level acquisition unit 12e acquires the level of the PWM control signal SW at edge timing. However, since the signal level EdgLev only needs to be acquired when the output pulse width correction is performed, the signal level acquisition unit 12e may acquire the level of the PWM control signal SW when the output pulse width correction unit 12f performs the correction.

[0033] As described above, the PWM control device 100 in Embodiment 1 has two features in acquiring the output pulse width of the PWM control signal SW. The first is that the output pulse width acquisition means 12 measures the output pulse width by detecting the edges (rising and falling edges) of the output pulse of the PWM control signal SW, and corrects the output pulse width measurement value Wmon if there is no output pulse of the PWM control signal SW.

[0034] The second point is that, in the correction of the output pulse width measurement value Wmon, the output pulse width measurement value Wmon is corrected to the PWM pulse period or zero according to the level of the PWM control signal SW ("high" or "low") at the time of edge detection (edge ​​timing) of the PWM control signal SW. As a result, when there is no output pulse from the PWM control signal SW, the output pulse width Wact will also be the PWM pulse period or zero, according to the level of the PWM control signal SW at the time of edge detection.

[0035] The operations described above, from edge detection to correction of the output pulse width measurement value Wmon, can be performed in the same way as in normal PWM control. In other words, there is no need to provide additional dedicated circuitry to realize the output pulse width acquisition means 12.

[0036] <Explanation of the effect of correcting the output pulse width measurement> The correction of the output pulse width measurement value Wmon by the output pulse width correction unit 12f will be explained in detail. First, as a comparative example, the case without correction will be explained using Figure 5. In Figure 5, the horizontal axis of the graph represents time. The vertical axis of the upper graph represents the level of the PWM control signal SW and the triangular wave (carrier) for PWM generation. The PWM control signal SW is "high" when it is "1" and "low" when it is "0". The vertical axis of the lower graph represents the output pulse width measurement value Wmon.

[0037] In Embodiment 1, the output pulse width measurement value Wmon is updated when the triangular wave used for PWM generation reaches a trough, and is updated to the final output pulse width at that point in time. For example, around time 0.0164, the output pulse width is T1, so at time 0.0165, the output pulse width measurement value Wmon is updated to T1. Also, at time 0.0166, the output pulse width of the falling edge output pulse is T2, so at time 0.0167, the output pulse width measurement value Wmon is updated to T2.

[0038] However, in the case of no output pulses, as shown in region A enclosed by the dashed line in Figure 5, the output pulse width measurement value Wmon is not updated. This means that when the duty cycle command Dref is 0% or 100%, the edges of the output pulse are not detected, and the output pulse width measurement value Wmon is not updated. If the output pulse width measurement value Wmon is not updated, it is not possible to obtain the current correct output pulse width Wact, and therefore, abnormality detection by comparing the command pulse width Wref and the output pulse width Wact cannot be performed correctly. This is also true when considering the output pulse width converted to the duty cycle output. In other words, if the output pulse width measurement value Wmon is not updated despite the duty cycle command Dref being 0% or 100%, a discrepancy will occur between the duty cycle command and the duty cycle output, even though the PWM control is normal, making it impossible to properly detect abnormalities.

[0039] Figure 6A shows the relationship between the PWM control signal and the output pulse width measurement value according to Embodiment 1, and illustrates the case where the output pulse width measurement value is corrected. Figure 6B is a diagram that explains the determination time for determining the absence of an output pulse in Embodiment 1. In Figure 6A, as in the example in Figure 5, the output pulse width measurement value Wmon is updated to T2 at time 0.0167 seconds. Also, similarly, there are no output pulses during the period included in region A. On the other hand, in Figure 6A, the signal level (low) of the PWM control signal SW at the falling edge F at time 0.0166 seconds is acquired, and the output pulse width measurement value Wmon is corrected accordingly. In the lower graph of Figure 6A, the output pulse width measurement value Wmon before correction (solid line) and the output pulse width measurement value Wmon after correction (dashed line) are equal until time 0.0168 seconds, but correction is performed at time 0.0168 seconds, and the output pulse width measurement value Wmon after correction becomes zero.

[0040] The determination of the absence of an output pulse is explained in Figure 6B. In Figure 6B, the threshold th for determining the absence of an output pulse is set to 100 microseconds, the same as the PWM pulse period. When the elapsed time since the last edge detection exceeds the threshold th, it is determined that there is no output pulse. The elapsed time since the last edge detection is obtained by measuring the time count value, as shown in the lower graph of Figure 6B. That is, the elapsed time is obtained from the difference dc (not shown) between the time count value at the current time Pt ​​and the time count value at the last edge detection time Ft. As can be seen from the upper graph of Figure 6B, the conversion from the time count value to the elapsed time (updating the elapsed time) is performed every 0.0001 seconds. In Figures 6A and 6B, the last edge detection time Ft is 0.0166 seconds, so the elapsed time exceeds the threshold th at time 0.0168 seconds. Therefore, it is determined that there is "no output pulse" at time 0.0168 seconds, and the output pulse width measurement value Wmon described above is corrected.

[0041] As described above, the absence of an output pulse is determined based on the elapsed time since the last edge detection, and the output pulse width measurement value Wmon is corrected to zero based on the level (low) of the PWM control signal SW at the time of the last edge detection, thereby accurately obtaining the output pulse width Wact. The same applies when the level of the PWM control signal SW at the time of the last edge detection is "high," in which case the output pulse width measurement value Wmon is corrected to the PWM pulse period. This means that even when the duty cycle command is 0% or 100% and no edge of the output pulse is detected, the output pulse width Wact of the PWM control signal SW can be accurately obtained, and anomaly detection can be correctly performed by comparing the command pulse width Wref and the output pulse width Wact.

[0042] In the explanations of Figures 5 and 6A, the update timing of the output pulse width measurement value Wmon is set to the timing of the trough of the triangular wave used for PWM generation, but the update timing is not limited to this. However, if the output pulse width measurement value Wmon is updated at the timing of the peak or trough of the triangular wave used for PWM generation, the update timing of the output pulse width measurement value Wmon will be the same as the update timing of general PWM control, which has the advantage of not requiring a separate timer for updating the output pulse width measurement value Wmon.

[0043] <Hardware configuration of the control unit> Next, the hardware configuration for realizing the control unit according to Embodiment 1 will be described. Figure 7 is a diagram showing an example of the hardware configuration for realizing the control unit according to Embodiment 1. Each function of the control unit 10 is realized by the processing circuits provided in the control unit 10. Specifically, the control unit 10 includes an arithmetic processing unit 81 such as a CPU, a storage device 82 that stores data from the arithmetic processing unit 81 and allows the arithmetic processing unit 81 to read the stored data, an input circuit 83 that inputs signals from external devices to the arithmetic processing unit 81, and an output circuit 84 that outputs signals from the arithmetic processing unit 81 to external devices. Data and signals are transmitted between the arithmetic processing unit 81, the storage device 82, the input circuit 83, and the output circuit 84 via a bus 85. Although not shown in the diagram, when the control unit 10 is configured as a microcontroller as described above, the control unit 10 also includes a timer.

[0044] As the arithmetic processing unit 81, for example, an ASIC (Application Specific Integrated Circuit), an IC (Integrated Circuit), a DSP (Digital Signal Processor), or an FPGA (Field Programmable Gate Array) can be used. Alternatively, an arithmetic processing circuit composed of various logic circuits and signal processing circuits may be used. Furthermore, multiple arithmetic processing units or arithmetic processing circuits mentioned above may be combined, with each arithmetic processing unit and arithmetic processing circuit performing a portion of the processing.

[0045] The storage device 82 includes RAM (Random Access Memory) configured to allow reading and writing of data from the arithmetic processing unit 81, ROM (Read Only Memory) configured to allow reading of data from the arithmetic processing unit 81, and the like.

[0046] The input circuit 83 is connected to a sensor 71 and a switch 72 that monitor the state of the controlled device 900, and inputs the output signals of the sensor 71 and the switch 72 to the arithmetic processing unit 81. The input circuit 83 includes an AD conversion unit, an interface circuit of the input circuit, etc. (all are omitted in the figure).

[0047] The output circuit 84 is connected to the controlled device 900, and includes an interface circuit etc. (omitted in the figure) that converts and outputs the output signal from the arithmetic processing unit 81.

[0048] Each function included in the control unit 10 is realized by the arithmetic processing unit 81 executing a software program stored in a storage device 82 such as a ROM, and cooperating with other hardware of the control unit 10 such as the storage device 82, the input circuit 83, and the output circuit 84. Note that setting data such as threshold values used by the control unit 10 is stored in the storage device 82 such as a ROM as part of the software program or as data separate from the software program.

[0049] [[ID=^{11}]] Each function mounted inside the control unit 10 may be composed of software modules, or may be composed of a combination of software and hardware.

[0050] <Operation of the PWM control device> Next, the operation will be described. FIG. 8 is a flowchart showing the operation of the PWM control device in Embodiment 1. First, a duty command Dref as a control command is generated (step ST101: control command generation step). As described above, the duty command Dref is generated based on a control command from a higher-level controller, an input from a user, or feedback from the controlled device 900.

[0051] Next, a PWM control signal SW is generated according to the duty command Dref (step ST102: control signal generation step). As described above, the PWM control signal SW is generated by comparing the duty command Dref with a carrier wave.

[0052] Next, the output pulse width Wact of the PWM control signal SW is acquired (Step ST103: Output pulse width acquisition step). As described above, the output pulse width Wact is determined based on the output pulse width measurement value Wmon, which is measured by detecting the edges of the PWM control signal SW. Also, as described above, the output pulse width measurement value Wmon is corrected depending on the presence or absence of output pulses. Details of the output pulse width acquisition step will be described later.

[0053] Next, the command pulse width Wref is calculated based on the duty cycle command Dref (Step ST104: Command pulse width calculation step). Note that the command pulse width Wref can be calculated if the duty cycle command Dref is available, so the command pulse width calculation step can be performed at any time after the control command generation step and before the abnormality determination step described later.

[0054] Next, the output pulse width Wact and the command pulse width Wref are compared to determine whether or not there is an abnormality in the PWM control (Step ST105: Abnormality Determination Step). Details of the abnormality determination step will be described later.

[0055] Next, if an abnormality is detected in the abnormality detection process, the process proceeds to step ST107; if no abnormality is detected, the process ends (step ST106).

[0056] If an abnormality is detected during the abnormality detection process, the generation and output of the PWM control signal SW are stopped (Step ST107: Output Stop Instruction Process).

[0057] <Flowchart of the output pulse width acquisition process> The output pulse width acquisition process will now be explained. Figure 9 is a flowchart showing the output pulse width acquisition process according to Embodiment 1. First, the edges of the PWM control signal SW are detected (step ST1031: output pulse detection process).

[0058] Next, the time counter is activated at the edge timing to obtain the edge time EdgT (Step ST1032: Edge time acquisition step).

[0059] Next, the signal level EdgLev of the PWM control signal SW at edge timing is obtained as either "high" or "low" (Step ST1033: Signal level acquisition step).

[0060] Next, the counter is operated at the edge timing, and the output pulse width measurement value Wmon is measured from the count value from rising edge to falling edge (or from falling edge to rising edge) (Step ST1034: Output pulse width measurement process).

[0061] Next, the presence or absence of an output pulse for the PWM control signal SW is determined based on the elapsed time since the last edge timing (step ST1035: output pulse presence / absence determination step). More specifically, the elapsed time since the last edge timing, i.e., the last edge detection time, is measured from the difference between the current time and the edge time EdgT, and if the elapsed time since the last edge timing is equal to or greater than the PWM pulse period, it is determined that there is no output pulse.

[0062] If it is determined that there is no output pulse, the process proceeds to step ST1037. If it is determined that there is an output pulse, the measured output pulse width Wmon is used as the output pulse width Wact, and the process ends (step ST1036).

[0063] If it is determined that there is no output pulse, the output pulse width measurement value Wmon is corrected, and the corrected output pulse width measurement value Wmon is set as the output pulse width Wact (Step ST1037: Output pulse width correction step). More specifically, if the signal level EdgLev is "low", the output pulse width measurement value Wmon is corrected to zero, and the value of "output pulse width Wact = 0". If the signal level EdgLev is "high", the output pulse width measurement value Wmon is corrected to the same value as the PWM pulse period, and the value of "output pulse width Wact = PWM pulse period".

[0064] <Flowchart of the abnormality detection process> The abnormality detection process will now be explained. Figure 10 is a flowchart showing the abnormality detection process according to Embodiment 1. First, the difference WDiff between the output pulse width Wact and the command pulse width Wref is calculated, and the difference WDiff is compared with the threshold Dth (Step ST1051: Pulse width comparison process).

[0065] Next, if the difference WDiff is greater than or equal to the threshold Dth, it is determined that there is an abnormality; if the difference WDiff is less than the threshold Dth, it is determined that there is no abnormality (normal) (Step ST1052: Difference abnormality determination step).

[0066] According to Embodiment 1, abnormalities can be detected even when the duty cycle command is 0% or 100%, and abnormalities in PWM control that occur within the PWM control device can be detected. More specifically, the PWM control device includes an output pulse width acquisition means for acquiring the output pulse width of the PWM control signal, a command pulse width calculation means for calculating the command pulse width, which is the pulse width of the duty cycle command, and an abnormality determination means for determining whether or not there is an abnormality in PWM control by comparing the output pulse width and the command pulse width. The output pulse width acquisition means measures an output pulse width measurement value, which is the measured value of the width of the output pulse of the PWM control signal, based on the edge of the PWM control signal, and determines whether or not there is an output pulse of the PWM control signal. If it is determined that there is no output pulse, it corrects the output pulse width measurement value and uses the corrected output pulse width measurement value as the output pulse width. If there is a difference between the output pulse width and the command pulse width that is greater than a threshold, it indicates that some kind of abnormality has occurred within the PWM control device. Therefore, abnormalities in PWM control that occur within the PWM control device can be detected.

[0067] Furthermore, if it is determined that there is no output pulse when acquiring the output pulse width, the measured output pulse width value is corrected, and the corrected measured output pulse width value is used as the output pulse width. In this way, if there is an output pulse width, the measured output pulse width value based on the edge is used as the output pulse width, and if there is no output pulse, such as when the duty cycle command is 0% or 100%, the last measured output pulse width value is corrected and used as the output pulse width. This makes it possible to acquire the output pulse width even when the duty cycle command is 0% or 100%, and to detect anomalies by comparing it with the command pulse width.

[0068] Furthermore, there is no need to provide additional dedicated circuits or hardware for anomaly detection. More specifically, the anomaly detection means determines the presence or absence of an anomaly by comparing the widths of two pulses, and does not compare signal levels. Therefore, dedicated circuits for comparing signal levels are unnecessary.

[0069] In Embodiment 1, the abnormality of PWM control is determined by comparing the command pulse width and the output pulse width. However, the abnormality of PWM control may also be determined by converting the output pulse width to a duty cycle output according to the PWM pulse period and comparing the duty cycle command with the duty cycle output.

[0070] Embodiment 2. Next, Embodiment 2 will be described with reference to Figures 11 to 18. In Embodiment 2, the controlled device includes an electric motor, and the PWM control device is included in a power conversion device for operating the electric motor. That is, a PWM control device similar to that in Embodiment 1 is applied to the power conversion device for operating the electric motor. More specifically, the PWM control device of Embodiment 2 outputs a plurality of PWM control signals to control the electric motor, and includes a rotation state acquisition means for acquiring the rotation state of the electric motor. It performs abnormality and normal determination of the PWM control according to the rotation state of the electric motor, and sets a signal level to stop the output of the PWM control according to the control state of the plurality of PWM control signals when an abnormality is determined.

[0071] The configuration and operation of the PWM control device of Embodiment 2 will be described below, focusing on the differences from Embodiment 1. Parts identical or equivalent to those in Embodiment 1 are denoted by the same reference numerals.

[0072] <Configuration of a power converter> Figure 11 is a block diagram showing the configuration of a power converter according to Embodiment 2. The power converter 1000 is a power converter that supplies AC power to an electric motor 901 to operate the electric motor 901, and includes a PWM control device 200 that generates a PWM control signal, and a drive circuit 903 that generates a drive signal to drive a three-phase inverter circuit 902 from the PWM control signal supplied from the PWM control device 200. The power converter 1000 also includes a three-phase (U-phase, V-phase, W-phase) AC bus 904 that connects the inverter circuit 902 and the electric motor 901, and a current sensor 905 that detects the current flowing through the AC bus 904 and outputs current detection values ​​for each phase (U-phase current detection value Iu, V-phase current detection value Iv, W-phase current detection value Iw, hereinafter referred to as three-phase current detection values ​​Iu, Iv, Iw). Furthermore, the power converter 1000 includes a voltage sensor 906 that detects the DC voltage input from the DC power supply 51 to the DC bus (not shown) of the inverter circuit 902 and outputs it as the input voltage Vpn. The three-phase current detection values ​​Iu, Iv, Iw and the input voltage Vpn are input to the control unit 20 of the PWM control device 200. The power converter 1000 is intended for use in electric vehicles such as electric cars and plug-in hybrid vehicles, and is designed to drive the motor, which is the power source, with power from a high-voltage battery.

[0073] The controlled device 900 is controlled by the PWM control device 200, similar to Embodiment 1, and some of its components overlap with those of the power converter 1000. The controlled device 900 includes an electric motor 901, an inverter circuit 902, and a drive circuit 903. The electric motor 901 is a motor that rotates a load (not shown) and is capable of regenerating the rotational energy of the load as electrical energy, and includes, for example, a permanent magnet three-phase AC synchronous motor or a three-phase brushless motor. The electric motor 901 is provided with a rotation angle sensor 907 that detects the rotor rotation angle θm of the electric motor 901. The rotation angle sensor 907 detects the rotor rotation angle θm of the rotor (not shown) of the electric motor 901 using a resolver or encoder or the like. The rotor rotation angle θm detected by the rotation angle sensor 907 is output to the control unit 20 of the PWM control device 200. The AC bus 904, current sensor 905, and voltage sensor 906 described above are also included in the controlled device 900.

[0074] The electric motor 901 operates by being supplied with alternating current by the inverter circuit 902. Furthermore, in the regenerative state of the electric motor 901, the generated regenerative power is used to charge the DC power supply 51 via the inverter circuit 902. Note that the PWM control device 200 is not limited to electric motors; other devices may also be controlled.

[0075] The inverter circuit 902 is connected to the DC power supply 51, for example, by a positive-side DC bus and a negative-side DC bus (both not shown), receives driving power from the DC power supply 51, and supplies regenerative power to the DC power supply 51. The inverter circuit 902 is also connected to the motor 901 by an AC bus 904, supplies driving power to the motor 901, and receives regenerative power from the motor 901. Although not shown, the inverter circuit 902 is generally an inverter in which six commonly known switching elements are connected in a full-bridge configuration, and has switching elements on the upper side of the U phase, the upper side of the V phase, the upper side of the W phase, the lower side of the U phase, the lower side of the V phase, and the lower side of the W phase. Each switching element is a semiconductor switching element such as a MOSFET (Metal Oxide Semiconductor Field Effect Transistor) having a built-in diode between the source and drain, for example. Note that the type and number of semiconductor switching elements are not limited to this, and semiconductor switching elements such as IGBTs (Insulated Gate Bipolar Transistors) and SiC-MOSFETs may be used.

[0076] The drive circuit 903 receives a PWM control signal from the control unit 20 of the PWM control device 200 and generates six drive signals (not shown) corresponding to the PWM control signal. Each switching element of the inverter circuit 902 is switched on and off by the drive signal generated by the drive circuit 903. The inverter circuit 902 converts the DC power supplied from the DC power supply 51 into AC by switching on and off each switching element, and supplies three-phase AC current to the motor 901.

[0077] <Configuration of PWM Control Device> The PWM control device 200 includes a control unit 20, which transmits the PWM control signals it generates to the controlled device 900. The control unit 20 generates PWM control signals corresponding to each switching element of the inverter circuit 902, namely, a PWM control signal SW_UH corresponding to the upper U-phase switching element, a PWM control signal SW_VH corresponding to the upper V-phase switching element, a PWM control signal SW_WH corresponding to the upper W-phase switching element, a PWM control signal SW_UL corresponding to the lower U-phase switching element, a PWM control signal SW_VL corresponding to the lower V-phase switching element, and a PWM control signal SW_WL corresponding to the lower W-phase switching element, and outputs them to the drive circuit 903.

[0078] Hereafter, the six PWM control signals mentioned above will be collectively referred to as PWM control signal SW**. Furthermore, the designation "_UH" indicates that it corresponds to the switching element on the upper U-phase of the inverter circuit 902, or to the PWM control signal that controls the switching element on the upper U-phase. Similarly, the designation "_VH" corresponds to the upper V-phase, "_WH" to the upper W-phase, "_UL" to the lower U-phase, "_VL" to the lower V-phase, and "_WL" to the lower W-phase.

[0079] <Functional block of the control unit> Figure 12 is a block diagram showing the configuration of the control unit according to Embodiment 2. The control unit 20 generates three-phase duty commands as control commands, namely U-phase duty command Dref_u, V-phase duty command Dref_v, and W-phase duty command Dref_w (hereinafter referred to as three-phase duty commands Dref_u, Dref_v, and Dref_w), and outputs the three-phase duty commands Dref_u, Dref_v, and Dref_w to the command pulse width calculation means 23 and the control signal generation means 29. The control unit 20 generates a PWM control signal SW** based on the three-phase duty commands Dref_u, Dref_v, and Dref_w, and outputs the PWM control signal SW** to the controlled device 900 (not shown in Figure 12) and the output pulse width acquisition means 22. The system includes a control signal generation means 29 for outputting signals, an output pulse width acquisition means 22 that acquires the output pulse widths Wact** (Wact_UH, Wact_VH, Wact_WH, Wact_UL, Wact_VL, Wact_WL), which are the pulse widths of each PWM control signal SW**, and outputs the output pulse widths Wact** to an abnormality determination means 27, and a command pulse width calculation means 23 that calculates the command pulse widths Wref** (Wref_UH, Wref_VH, Wref_WH, Wref_UL, Wref_VL, Wref_WL) based on the three-phase duty commands Dref_u, Dref_v, and Dref_w, and outputs the command pulse widths Wref** to an abnormality determination means 27.

[0080] The control unit 20 further includes an abnormality determination means 27 that determines whether or not there is an abnormality in the PWM control based on the comparison result of each output pulse width Wact** and each command pulse width Wref**, and outputs the result as a state determination result ErrSt** (ErrSt_UH, ErrSt_VH, ErrSt_WH, ErrSt_UL, ErrSt_VL, ErrSt_WL), and an output stop instruction means 18 that receives the state determination result ErrSt** from the abnormality determination means 27 and outputs a stop instruction signal Stp to the control signal generation means 29 according to the state determination result ErrSt**. The abnormality determination means 27 determines whether or not there is an abnormality for each of the U-phase upper side, etc., by performing abnormality determination by comparing the output pulse width Wact** and the command pulse width Wref** for each of the U-phase upper side, etc.

[0081] Furthermore, the control unit 20 includes a rotation state acquisition means 25 that calculates the electrical angle θe and electrical angular period Tω based on the rotor rotation angle θm, outputs the electrical angle θe to the control command generation means 21, and outputs the electrical angular period Tω to the normal control determination means 26, and a normal control determination means 26 that determines whether the PWM control is operating normally based on the electrical angular period Tω and the state determination result ErrSt**.

[0082] The control command generation means 21 will now be described. Figure 13 is a block diagram showing the configuration of the control command generation means according to Embodiment 2. Examples of control commands for controlling the electric motor 901 include torque commands, current commands, voltage commands, etc., but in Embodiment 2, a torque command Trq* is used as a control command as an example. The control command generation means 21 includes a current command generation unit 21a that generates current commands in the dq-axis coordinate system, i.e., d-axis current command Id* and q-axis current command Iq*, based on the torque command Trq*; a three-phase to two-phase conversion unit 21c that converts three-phase current detection values ​​Iu, Iv, and Iw into current detection values ​​in the dq-axis coordinate system, i.e., d-axis current detection value Id and q-axis current detection value Iq; and a d-axis current generation unit 21c that converts three-phase current detection values ​​Iu, Iv, and Iw into current detection values ​​in the dq-axis coordinate system, i.e., d-axis current detection value Id and q-axis current detection value Iq, respectively, based on the deviation ΔId (not shown) between the d-axis current command Id* and the d-axis current detection value Id, and the deviation ΔIq (not shown) between the q-axis current command Iq* and the q-axis current detection value Iq, respectively, The system includes a voltage command generation unit 21b that generates standard voltage commands, namely the d-axis voltage command Vdc and the q-axis voltage command Vqc; a two-phase to three-phase conversion unit 21d that converts the d-axis voltage command Vdc and the q-axis voltage command Vqc into three-phase voltage commands, namely the U-phase voltage command Vuc, the V-phase voltage command Vvc, and the W-phase voltage command Vwc (hereinafter referred to as three-phase voltage commands Vuc, Vvc, and Vwc); and a duty cycle command calculation unit 21e that calculates three-phase duty cycle commands Dref_u, Dref_v, and Dref_w based on the three-phase voltage commands Vuc, Vvc, and Vwc.

[0083] The current command generation unit 21a receives a torque command Trq* from a higher-level controller or system (not shown) and generates a d-axis current command Id* and a q-axis current command Iq* based on the torque command Trq*. In the dq-axis coordinate system, the d-axis indicates the magnetic pole position of the electric motor 901, i.e., the direction of the magnetic flux, and the q-axis indicates the direction electrically perpendicular to the d-axis. The dq-axis coordinate system is a rotating coordinate system, and when the rotor of the electric motor 901, which has magnets, rotates, the dq-axis coordinate system also rotates.

[0084] The voltage command generation unit 21b performs current feedback calculations for current commands in the dq-axis coordinate system, and generates d-axis voltage commands Vdc and q-axis voltage commands Vqc such that the deviation ΔId for the d-axis current and the deviation ΔIq for the q-axis current converge to zero, respectively.

[0085] The three-phase to two-phase conversion unit 21c receives three-phase current detection values ​​Iu, Iv, Iw and electrical angle θe from the current sensor 905 and the rotation state acquisition means 25, and uses the electrical angle θe to convert the three-phase current detection values ​​Iu, Iv, and Iw into d-axis current detection value Id and q-axis current detection value Iq, which are current detection values ​​in the dq-axis coordinate system.

[0086] The two-phase to three-phase conversion unit 21d receives the d-axis voltage command Vdc and the q-axis voltage command Vqc from the voltage command generation unit 21b, as well as the electrical angle θe from the rotation state acquisition means 25. Using the electrical angle θe, the unit converts the d-axis voltage command Vdc and the q-axis voltage command Vqc into three-phase voltage commands Vuc, Vvc, and Vwc. Preferably, the three-phase voltage commands Vuc, Vvc, and Vwc are set to be less than or equal to the input voltage Vpn input to the inverter circuit 902.

[0087] The duty cycle command calculation unit 21e receives three-phase voltage commands Vuc, Vvc, and Vwc from the two-phase to three-phase conversion unit 21d, as well as an input voltage Vpn from the voltage sensor 906. From the three-phase voltage commands Vuc, Vvc, and Vwc and the input voltage Vpn, it generates three-phase duty cycle commands Dref_u, Dref_v, and Dref_w. At this time, the duty cycle command calculation unit 21e generates three-phase duty cycle commands Dref_u, Dref_v, and Dref_w corresponding to the optimal correction control command. The duty cycle command calculation unit 21e outputs the generated three-phase duty cycle commands Dref_u, Dref_v, and Dref_w to the control signal generation means 29.

[0088] The output pulse width acquisition means 22 acquires the output pulse width Wact**. As described above, the output pulse width Wact** includes six output pulse widths. The acquisition of each output pulse width is the same as that of the output pulse width acquisition means 12 in Embodiment 1.

[0089] The command pulse width calculation means 23 calculates the command pulse width Wref**. As described above, the command pulse width Wref** includes six command pulse widths. The calculation of each command pulse width is the same as that of the command pulse width calculation means 13 in Embodiment 1.

[0090] The rotation state acquisition means 25 will now be described. Figure 14 is a block diagram showing the configuration of the rotation state acquisition means according to Embodiment 2. The rotation state acquisition means 25 includes an electrical angle calculation unit 25a that receives the rotor rotation angle θm of the electric motor 901 from the rotation angle sensor 907 and converts the rotor rotation angle θm to an electrical angle θe based on the number of pole pairs of the electric motor 901, and an electrical angle period calculation unit 25b that calculates the electrical angle period Tω from the electrical angle θe. The electrical angle calculation unit 25a outputs the calculated electrical angle θe to the control command generation means 21 and the electrical angle period calculation unit 25b. The electrical angle period calculation unit 25b outputs the calculated electrical angle period Tω to the abnormality determination means 27 and the normal control determination means 26. The electrical angle period Tω is an example of a time determined based on the rotation state of the electric motor 901.

[0091] The normal control determination means 26 determines, based on the determination result by the abnormality determination means 27 and the rotation state of the electric motor 901 acquired by the rotation state acquisition means 25, that the PWM control signal SW** is operating normally without sticking. Here, "sticking" refers to the PWM control signal SW** continuing to maintain a "high" or "low" state in an unintended way, indicating an abnormal state in which the signal level of the PWM control signal SW** cannot be switched normally. Sticking of the PWM control signal SW** can occur when input from sensors, etc., cannot be acquired due to an abnormality in the input circuit of the PWM control device 200, when memory rewriting is not performed normally, or when the arithmetic unit is not operating normally. In addition, if the control unit 20 is configured as a microcomputer, sticking of the PWM control signal SW** can also occur if there is an abnormality in the output circuit of the microcomputer (such as when the register value of the output timer used for PWM control is not updated normally).

[0092] The normal control determination means 26 determines that the PWM control is normal if the state determination result ErrSt** from the abnormality determination means 27, which will be described later, remains "normal" for half the electrical angular period Tω (hereinafter referred to as "Tω / 2"). Details of the PWM control normal determination will be described later.

[0093] The abnormality determination means 27 will now be described. Figure 15 is a block diagram showing the configuration of the abnormality determination means according to Embodiment 2. The abnormality determination means 27 includes a pulse width comparison unit 27a that calculates the difference WDiff** (WDiff_UH, WDiff_VH, WDiff_WH, WDiff_UL, WDiff_VL, WDiff_WL) between the command pulse width Wref** and the output pulse width Wact**, a difference abnormality determination unit 27b that determines whether or not there is an abnormality in the difference WDiff** and outputs the result as a difference abnormality determination result ErrD** (ErrD_UH, ErrD_VH, ErrD_WH, ErrD_UL, ErrD_VL, ErrD_WL), and a pulse width fixing time FixT** (FixT_UH, FixT_VH, FixT_WH, FixT_UL, FixT_VL, F The system includes a pulse width fixing time measurement unit 27c that measures as ixT_WL), a rotational fluctuation determination unit 27d that determines whether or not there is a time fluctuation in the ideal value of the PWM control command, a control command abnormality determination unit 27e that determines whether or not there is an abnormality in the PWM control command and outputs the result as a control command abnormality determination result ErrR** (ErrR_UH, ErrR_VH, ErrR_WH, ErrR_UL, ErrR_VL, ErrR_WL), and an abnormality presence / absence determination unit 27f that determines whether or not there is an abnormality based on the differential abnormality determination result ErrD** and the control command abnormality determination result ErrR** and outputs the result as an abnormality determination result Err** (Err_UH, Err_VH, Err_WH, Err_UL, Err_VL, Err_WL).

[0094] The abnormality determination means 27 includes an on-fixation determination unit 27g that determines whether the output pulse of the PWM control signal SW** is on-fixed and outputs the result as an on-fixation determination result FixON** (FixON_UH, FixON_VH, FixON_WH, FixON_UL, FixON_VL, FixON_WL), an off-fixation determination unit 27h that determines whether the output pulse of the PWM control signal SW** is off-fixed and outputs the result as an off-fixation determination result FixOFF** (FixOFF_UH, FixOFF_VH, FixOFF_WH, FixOFF_UL, FixOFF_VL, FixOFF_WL), and a control state determination unit 27i that determines the state of PWM control based on the abnormality determination result Err**, the on-fixation determination result FixON**, and the off-fixation determination result FixOFF** and outputs the result as a state determination result ErrSt**.

[0095] The pulse width comparison unit 27a receives the command pulse width Wref** and the output pulse width Wact** from the command pulse width calculation means 23 and the output pulse width acquisition means 22, respectively. As described above, the command pulse width Wref** and the output pulse width Wact** include six command pulse widths and output pulse widths corresponding to each switching element of the inverter circuit 902, such as the U-phase upper side. The pulse width comparison unit 27a calculates the difference between the corresponding command pulse width and the output pulse width, and calculates the difference WDiff**. The pulse width comparison unit 27a outputs the difference WDiff** to the difference anomaly determination unit 27b.

[0096] The difference anomaly determination unit 27b determines whether each of the six differences included in the difference WDiff** is greater than or equal to the threshold Dth. As a result, the difference anomaly determination result ErrD** indicates whether or not there is a difference anomaly for each switching element, such as the upper U-phase. The difference anomaly determination unit 27b outputs the difference anomaly determination result ErrD** to the anomaly presence / absence determination unit 27f.

[0097] The pulse width fixation time measurement unit 27c receives the output pulse width Wact** from the output pulse width acquisition means 22 and measures the time during which the same value persists for each output pulse width included in the output pulse width Wact** as the pulse width fixation time. The pulse width fixation time measurement unit 27c outputs the pulse width fixation time FixT** to the control command abnormality determination unit 27e and the on-fixation determination unit 27g. Note that if the PWM pulse period is varied, the upper limit of the output pulse width will change, so the pulse width fixation time measurement unit 27c also measures the time during which the upper limit of the output pulse width persists as the fixation time.

[0098] The rotational fluctuation determination unit 27d determines whether or not there is a time variation in the ideal value of the PWM control command (three-phase duty cycle commands Dref_u, Dref_v, Dref_w) to be output. The rotational fluctuation determination unit 27d receives a torque command Trq* from a higher-level controller or system (not shown) and determines "no time variation" if the torque command Trq* is zero, and determines "time variation exists" if the torque command Trq* is not zero. The rotational fluctuation determination unit 27d outputs the above determination result as time variation presence / absence Ref_v to the control command abnormality determination unit 27e. In Embodiment 2, the presence or absence of a time variation in the ideal value of the PWM control command is determined according to the torque command Trq*, but it is also possible to configure the unit to determine whether or not there is a time variation in the ideal value of the PWM control signal SW** according to the duty cycle command (three-phase duty cycle commands Dref_u, Dref_v, Dref_w). For example, it is possible to determine "no time variation" when the duty cycle command is 50%.

[0099] The control command abnormality determination unit 27e receives the pulse width fixing time FixT** and electrical angle period Tω from the pulse width fixing time measurement unit 27c and the rotation state acquisition means 25, and the presence or absence of time variation Ref_v from the rotation variation determination unit 27d. The control command abnormality determination unit 27e determines that there is a "control command abnormality" if the presence or absence of time variation Ref_v is "there is a time variation" and the pulse width fixing time FixT** is Tω / 2 or greater. If the above conditions are not met, that is, if there is no time variation Ref_v or the pulse width fixing time FixT** is less than Tω / 2, it determines that there is "no control command abnormality".

[0100] Furthermore, the control command abnormality determination unit 27e may not perform a determination if the rotational speed of the motor 901 is below a predetermined rotational speed. In this case, the control command abnormality determination unit 27e will only determine a control command abnormality if the rotational speed of the motor 901 is greater than the predetermined rotational speed. The values ​​of the three-phase duty commands Dref_u, Dref_v, and Dref_w are determined according to the electrical angle θe, but the electrical angle θe changes with the rotation of the motor 901. Therefore, if the motor 901 is not rotating, the electrical angle θe becomes a fixed value, and even if the PWM control is normal, the values ​​of the three-phase duty commands Dref_u, Dref_v, and Dref_w may remain the same. By determining a control command abnormality only when the rotational speed of the motor 901 is greater than a predetermined rotational speed, as described above, it is possible to prevent false detection of abnormalities when the motor 901 is not rotating. The rotational speed of the electric motor 901 can be obtained by calculating the time change of the rotor rotation angle θm using the rotational state acquisition means 25.

[0101] Furthermore, while the control command abnormality determination unit 27e in Embodiment 2 determines whether or not there is a control command abnormality using the fixation time of the output pulse width Wact**, the presence or absence of a control command abnormality may also be determined using the fixation time of the command pulse width Wref**. In this case, the command pulse width Wref** is input to the pulse width fixation time measurement unit 27c, and the pulse width fixation time measurement unit 27c measures the time during which the same value persists for each command pulse width included in the command pulse width Wref** as the pulse width fixation time.

[0102] Furthermore, while the system currently determines a "control command anomaly" when the pulse width fixation time (FixT**) is greater than or equal to Tω / 2, a time smaller than Tω / 2 may be used as the determination value for a control command anomaly. In this case, the control command anomaly can be detected earlier.

[0103] The abnormality detection unit 27f receives the differential abnormality detection result ErrD** and the control command abnormality detection result ErrR** from the differential abnormality detection unit 27b and the control command abnormality detection unit 27e, respectively. If at least one of the differential abnormality detection result ErrD** and the control command abnormality detection result ErrR** indicates an abnormality, the unit determines that there is no abnormality if neither indicates an abnormality. The unit outputs the result as the abnormality detection result Err** to the control state determination unit 27i. Both the differential abnormality detection result ErrD** and the control command abnormality detection result ErrR** contain six determination results corresponding to each switching element of the inverter circuit 902. Therefore, the abnormality detection unit 27f also determines the presence or absence of an abnormality for each of these results. For example, if the U-phase upper side has "Differential Anomaly Judgment Result: Anomaly Found" and "Control Command Anomaly Judgment Result: No Anomaly," and the V-phase upper side has "Differential Anomaly Judgment Result: No Anomaly" and "Control Command Anomaly Judgment Result: No Anomaly," then in the Anomaly Judgment Result Err**, the U-phase upper side will have "Anomaly Found," and the V-phase upper side will have "No Anomaly."

[0104] As described above, by combining abnormality detection based on the difference between the command pulse width and the output pulse width with abnormality detection regarding the control command, it is possible to detect abnormalities in PWM control even when memory rewriting is not performed correctly within the control unit 20, or when the arithmetic unit is not operating correctly.

[0105] The ON-lock determination unit 27g receives the output pulse width Wact** and pulse width lock time FixT** from the output pulse width acquisition means 22 and the pulse width lock time measurement unit 27c, respectively, and determines whether or not ON-locking has occurred for each switching element of the inverter circuit 902 based on the output pulse width Wact** and pulse width lock time FixT**, and outputs the result as the ON-lock determination result FixON** to the control state determination unit 27i. The ON-lock determination unit 27g determines that ON-locking has occurred (the PWM control signal is maintaining a high state in an unintended way) if the pulse width lock time FixT** is greater than or equal to a predetermined time (for example, Tω / 2 or more) and the output pulse width Wact** is the PWM pulse period. As described above, the determination of whether or not ON-locking has occurred is performed for each switching element of the inverter circuit 902. For this reason, the ON-lock determination result FixON** includes six ON-lock determination results, such as the upper U-phase.

[0106] The off-lock determination unit 27h receives the output pulse width Wact** and pulse width lock time FixT** from the output pulse width acquisition means 22 and the pulse width lock time measurement unit 27c, respectively. Based on the output pulse width Wact** and pulse width lock time FixT**, it determines whether or not off-locking has occurred for each switching element of the inverter circuit 902, and outputs the result as the off-lock determination result FixOFF** to the control state determination unit 27i. The off-lock determination unit 27h determines that off-locking has occurred (the PWM control signal is maintaining a low state in an unintended way) if the pulse width lock time FixT** is greater than or equal to a predetermined time (for example, Tω / 2 or more) and the output pulse width Wact** is zero. As described above, the determination of whether or not off-locking has occurred is performed for each switching element of the inverter circuit 902. For this reason, the off-lock determination result FixOFF** includes six off-lock determination results, such as the upper U-phase.

[0107] The control state determination unit 27i receives the abnormality determination result Err** from the abnormality presence / absence determination unit 27f, and the ON-fixed determination result FixON** and OFF-fixed determination result FixOFF** from the ON-fixed determination unit 27g and OFF-fixed determination unit 27h, respectively. Based on the abnormality determination result Err**, the ON-fixed determination result FixON**, and the OFF-fixed determination result FixOFF**, the control state determination unit 27i determines the state of the PWM control and outputs the state determination result ErrSt** to the output stop instruction means 28 and the normal control determination means 26.

[0108] The status determination result ErrSt** has four possible outcomes: "Normal", "ON-lock abnormal", "OFF-lock abnormal", and "Pulse abnormal". The control status determination unit 27i sets the status determination result ErrSt** to "Normal" if the abnormality determination result Err** is "No abnormality". If the abnormality determination result Err** is "Abnormal", and the ON-lock determination result FixON** is "ON-locked", the status determination result ErrSt** is set to "ON-lock abnormal". If the abnormality determination result Err** is "Abnormal", and the OFF-lock determination result FixOFF** is "OFF-locked", the status determination result ErrSt** is set to "OFF-lock abnormal". Furthermore, if the abnormality determination result Err** is "Abnormal", and the ON-lock determination result FixON** and OFF-lock determination result FixOFF** are "ON-not locked" and "OFF-not locked", respectively, the control status determination unit 27i sets the status determination result ErrSt** to "Pulse abnormal". As described above, "on-lock abnormality" and "off-lock abnormality" indicate the type of abnormality when the abnormality determination means 27 determines that there is an abnormality in the PWM control. The determination of whether or not there is on-lock and off-lock may be performed separately from the determination of whether or not there is an abnormality in the PWM control. For this reason, the on-lock determination unit 27g and the off-lock determination unit 27h do not necessarily need to be provided in the abnormality determination means 27. Also, the pulse width lock time measurement unit 27c measures the pulse width lock time of the output pulse width Wact** and does not perform abnormality determination itself, so it does not necessarily need to be provided in the abnormality determination means 27.

[0109] Furthermore, the state determination result ErrSt** also includes six state determination results corresponding to each switching element of the inverter circuit 902, such as the upper U-phase.

[0110] The output stop instruction means 28, similar to the output stop instruction means 18 in Embodiment 1, stops the generation of the PWM control signal SW** by the control signal generation means 29 according to the state of the PWM control indicated by the state determination result ErrSt**. As described above, the state determination result ErrSt** includes six state determination results corresponding to each switching element of the inverter circuit 902, such as the state determination result for the upper U-phase. Therefore, the output stop instruction means 28 generates a stop instruction signal Stp according to the state of each PWM control signal corresponding to each switching element.

[0111] Furthermore, in Embodiment 2, the PWM control device 200 is applied to the power conversion device 1000 for operating the electric motor 901, and a stop instruction signal Stp is generated that sets the level of the PWM control signal according to the control state of each of the multiple PWM control signal SW**. For example, in order to stop the phase current of the electric motor 901 within the inverter circuit 902 and prevent power from being regenerated to the input side of the inverter circuit 902, the output of the PWM control signal SW** may be stopped by turning on all or all of the upper switching elements of the inverter circuit 902, thereby short-circuiting each phase of the electric motor 901 with each other, resulting in a so-called three-phase short circuit state.

[0112] For example, if the state determination result ErrSt_UH for the U-phase upper PWM control signal SW_UH is "off-fixed abnormality", even if the PWM control signal SW** is instructed to stop outputting to turn on all the upper switching elements of the inverter circuit 902 and turn off all the lower switching elements to short-circuit the three phases, the U-phase upper switching elements remain off, and therefore the three phases cannot be short-circuited. For this reason, the output stop instruction means 28 generates a stop instruction signal Stp corresponding to the normal or abnormal state of the multiple PWM control signal SW** and outputs it to the control signal generation means 29.

[0113] If all status determination results ErrSt** are "normal", the output stop instruction means 28 does not issue an output stop instruction.

[0114] If the state determination result ErrSt** for any of the upper switching elements (e.g., the upper U-phase) is "on-locked abnormal," the output stop instruction means 28 generates a stop instruction signal Stp that sets the signal levels of the PWM control signals SW_UH, SW_VH, and SW_WH for all of the upper switching elements to "high," and the signal levels of the PWM control signals SW_UL, SW_VL, and SW_WL for all of the lower switching elements to "low." The same applies if the state determination result ErrSt** for any of the lower switching elements is "on-locked abnormal."

[0115] If the state determination result ErrSt** for any of the upper switching elements (e.g., the upper U-phase) is "off-locked abnormality", the output stop instruction means 28 generates a stop instruction signal Stp that sets the signal levels of the PWM control signals SW_UH, SW_VH, and SW_WH for all of the upper switching elements to "low", and the signal levels of the PWM control signals SW_UL, SW_VL, and SW_WL for all of the lower switching elements to "high". The same applies if the state determination result ErrSt** for any of the lower switching elements is "off-locked abnormality".

[0116] As described above, in the case of an "on-locked abnormality," all PWM control signals for the switching elements on the same side (upper or lower) as the switching element determined to be "on-locked" are set to "high," and all PWM control signals for the switching elements on the opposite side (lower or upper) are set to "low." In the case of an "off-locked abnormality," all PWM control signals for the switching elements on the same side (upper or lower) as the switching element determined to be "off-locked" are set to "low," and all PWM control signals for the switching elements on the opposite side (lower or upper) are set to "high." In this way, the output stop instruction means 28 sets the state of the switching elements at the time of output stop according to the type of abnormality.

[0117] For a switching element whose state determination result ErrSt** is "pulse abnormality", the output stop instruction means 28 generates a stop instruction signal Stp to stop the output of the PWM control signal.

[0118] The control signal generation means 29 generates and outputs PWM control signals SW** for controlling the controlled device 900 according to the three-phase duty commands Dref_u, Dref_v, and Dref_w. The generation of PWM control signals SW by the control signal generation means 29 is the same as that of the control signal generation means 19 in Embodiment 1, and is performed by comparing the three-phase duty commands Dref_u, Dref_v, and Dref_w with the carrier. However, the control signal generation means 29 generates six PWM control signals SW** corresponding to each switching element of the inverter circuit 902. Furthermore, when the control signal generation means 29 receives a stop instruction signal Stp from the output stop instruction means 28, it stops the generation and output of some or all of the PWM control signals SW** according to the content of the stop instruction signal Stp. Alternatively, it sets the signal level to "high" or "low".

[0119] With the above configuration, when the electric motor 901 is the target of control, the output of the PWM control signals can be stopped according to the control status of the multiple PWM control signals SW**, and the electric motor 901 can be stopped early if an abnormality occurs in the PWM control.

[0120] Three features of the PWM control device 200 in Embodiment 2 will now be described. The first is the normal determination by the normal control determination means 26 described above. As described above, the PWM control is determined to be normal based on the abnormal determination result by the abnormal determination means 27 and the rotation state (electrical angular period Tω) of the electric motor 901 acquired by the rotation state acquisition means 25. This confirms that the PWM control is operating normally without the PWM control signal SW** becoming stuck.

[0121] The second point is that, even if there is no difference between the output pulse width Wact** and the command pulse width Wref**, the abnormality determination means 27 determines an abnormality in PWM control based on the rotation state (electrical angular period Tω) of the electric motor 901 and the output pulse width Wact** obtained by the rotation state acquisition means 25.

[0122] The third feature is that the abnormality determination means 27 determines on-fixation abnormalities and off-fixation abnormalities of the PWM control signal SW** by measuring the time during which the output pulse width remains constant, and the output stop instruction means 28 generates a stop instruction signal Stp according to the abnormal state of each PWM control signal SW** corresponding to each switching element.

[0123] In Embodiment 2, as in Embodiment 1, even when the duty cycle command is 0% or 100%, the accurate output pulse width Wact** can be obtained by correcting the output pulse width measurement value Wmon. Therefore, just like in Embodiment 1, abnormality detection can be correctly performed based on the comparison result (difference WDiff**) between the command pulse width Wref** and the output pulse width Wact**.

[0124] However, when the PWM duty cycle command is 0% or 100%, even if there is no difference between the command pulse width Wref** and the output pulse width Wact**, if, for example, the PWM control signal SW** is stuck as described above, it becomes uncertain whether the PWM control is actually operating according to the duty cycle command. The normal control determination means 26 of Embodiment 2 determines that normal PWM control is being performed even in such cases. This will be explained below.

[0125] Figure 16 is a diagram comparing the time waveform of the electrical angle and the time waveform of the PWM control signal on the upper U-phase in Embodiment 2, where the upper graph shows the time waveform of the electrical angle θe and the lower graph shows the PWM control signal SW_UH on the upper U-phase. As described above, the three-phase voltage commands Vuc, Vvc, and Vwc are calculated based on the d-axis voltage command Vdc and the q-axis voltage command Vqc and the electrical angle θe. Therefore, the three-phase duty cycle commands Dref_u, Dref_v, and Dref_w calculated based on the three-phase voltage commands Vuc, Vvc, and Vwc, and the PWM control signal SW** generated based on the three-phase duty cycle commands Dref_u, Dref_v, and Dref_w, fluctuate with an electrical angle period Tω. In such a case, if the PWM control is operating normally, the duty cycle command will not remain at 0% or 100% for Tω / 2 or longer. In other words, if the state determination result ErrSt**, which is the output of the abnormality determination means 27, remains in a "normal" state for Tω / 2 or longer, then the signal level of the PWM control signal SW** is "high" or The PWM control is functioning correctly without getting stuck in the "low" state. In Figure 16, there is a period T3 around time 0.017 seconds during which there is no output pulse for the PWM control signal, but this period T3 never lasts longer than Tω / 2. This is why the normal control determination means 26 determines that the PWM control is normal if the state determination result ErrSt** remains "normal" for Tω / 2 or longer.

[0126] The abnormality determination by the differential abnormality determination unit 27b, that is, the abnormality determination based on the comparison of the command pulse width Wref** and the output pulse width Wact**, assumes that the command pulse width Wref** is correct, and therefore cannot detect abnormalities that occurred before the generation of the three-phase duty commands Dref_u, Dref_v, and Dref_w, which are the basis for calculating the command pulse width Wref**. In the case where the PWM control device 200 is applied to the power conversion device 1000 for operating the electric motor 901, as in Embodiment 2, it is also necessary to detect abnormalities that occurred before the generation of the three-phase duty commands Dref_u, Dref_v, and Dref_w. In Embodiment 2, focusing on the fact that if PWM control is operating normally, the control commands (including the three-phase duty cycle commands Dref_u, Dref_v, and Dref_w) within the control unit 20 will not continue to have the same value for Tω / 2 or more, if the output pulse width Wact** of the PWM control signal SW**, which is the output of the control unit 20, continues to have the same value for Tω / 2 or more, the control command abnormality determination unit 27e, the on-fixation determination unit 27g, and the off-fixation determination unit 27h determine that an abnormality has occurred within the control unit 20, such as a halt in the calculation of the control command or memory fixation.

[0127] However, the time variation range of the three-phase duty cycle commands Dref_u, Dref_v, and Dref_w may be small. In Embodiment 2, the three-phase duty cycle commands Dref_u, Dref_v, and Dref_w are generated based on the torque command Trq* input to the control command generation means 21. For example, if the torque command is zero, the value of the generated duty cycle command will be around 50%, and this value may persist even if the PWM control is operating normally. In such cases, abnormality detection based on the pulse width fixing time FixT** of the output pulse width Wact** can be omitted. This prevents false detection of abnormalities.

[0128] As described above, the control command abnormality determination unit 27e of Embodiment 2 determines that there is a "control command abnormality" when the time variation presence / absence Ref_v is "time variation present" and the pulse width fixing time FixT** is Tω / 2 or greater. When the torque command Trq* is zero, the time variation presence / absence Ref_v becomes "no time variation", and regardless of the pulse width fixing time FixT**, there is "no control command abnormality". Therefore, false detection of abnormalities when the time variation range of the three-phase duty commands Dref_u, Dref_v, and Dref_w is small is prevented.

[0129] As described above, the PWM control device of Embodiment 2, when applied to a power conversion device for operating an electric motor, can confirm that the PWM control is operating normally without the PWM control signal becoming stuck. Furthermore, it can detect abnormalities in the PWM control even when abnormalities such as the control command calculation stopping or memory sticking occur. In addition, it can appropriately instruct the stopping of the output of the PWM control signal according to the control state of multiple PWM control signals, and stop the operation of the electric motor at the appropriate timing.

[0130] In Embodiment 2, the PWM control signal SW**, which is the output of the PWM control device 200, is monitored and the output pulse width of the PWM control signal SW** is acquired. However, it is also possible to monitor the drive signal, which is the output of the drive circuit 903, and acquire the output pulse width of the drive signal. In this case, the output pulse width of the drive signal and the command pulse width Wref** are compared in the abnormality determination, and abnormalities occurring in the transmission path from the PWM control device 200 to the motor 901 can also be detected. Furthermore, if the state buffer IC or gate driver that is the output destination of the microcomputer fails, the drive signal, which is the output of the drive circuit 903, may become stuck. For this reason, by monitoring the drive signal, which is the output of the drive circuit 903, and acquiring the output pulse width of the drive signal, it is possible to determine whether or not the drive signal is stuck.

[0131] In addition, although the power conversion device according to Embodiment 2 has been described assuming an inverter that converts DC power into AC power, the type of the power conversion device is not limited to this, and any power conversion device that includes semiconductor switching elements and converts the output form of power may be used. For example, it may be an AC / DC converter (Alternate Current / Direct Current Converter) that converts AC power into DC power, or a DC / DC converter (Direct Current / Direct Current Converter) that changes the levels of the voltage and current of DC power and outputs them.

[0132] <Operation of PWM Control Device> Next, the operation will be described. FIG. 17 is a flowchart showing the operation of the PWM control device in Embodiment 2. First, the rotation state of the motor 901 is acquired (step ST201: rotation state acquisition step). More specifically, the rotor rotation angle θm of the motor 901 is acquired from the rotation angle sensor 907, and the electrical angle θe and the electrical angle period Tω are calculated based on the rotor rotation angle θm.

[0133] Next, three-phase duty commands Dref_u, Dref_v, and Dref_w as control commands are generated (step ST202: control command generation step).

[0134] Next, six PWM control signals SW** are generated according to the three-phase duty commands Dref_u, Dref_v, and Dref_w (step ST203: control signal generation step).

[0135] Next, the output pulse width Wact** of each of the PWM control signals SW** is acquired (step ST204: output pulse width acquisition step). The acquisition of each output pulse width Wact** is the same as the output pulse width acquisition step described in Embodiment 1.

[0136] Next, six command pulse widths Wref** are calculated based on the three-phase duty commands Dref_u, Dref_v, and Dref_w (step ST205: command pulse width calculation step).

[0137] Next, the presence or absence of an abnormality in the PWM control is determined (Step ST206: Abnormality Determination Step). Details of the abnormality determination step will be described later.

[0138] Next, if an abnormality is detected in the abnormality detection process, that is, if any of the six state determination results ErrSt** is "on-lock abnormality", "off-lock abnormality", or "pulse abnormality", the process proceeds to step ST208. If no abnormality is detected, that is, if all six state determination results ErrSt** are "normal", the process proceeds to step ST209 (step ST207).

[0139] If an abnormality is detected in the abnormality detection process, the generation and output of the PWM control signal SW** are stopped (step ST208: output stop instruction process). As described above, in Embodiment 2, the PWM control device 200 is applied to the power converter 1000 for operating the electric motor 901, and the output of the PWM control signal SW** is stopped according to the type of abnormality and the switching element where the abnormality is occurring, such as stopping the output of the PWM control signal SW** in a three-phase short-circuit state to prevent power regeneration to the input side. Details regarding the stopping of the output of the PWM control signal SW** are as described above.

[0140] If no abnormality is detected in the abnormality detection process, the PWM control is determined to be operating normally based on the rotational state (electrical angular period Tω) of the electric motor 901 (step ST209: normality determination process). More specifically, the PWM control is determined to be operating normally if all six state determination results ErrSt** remain "normal" for Tω / 2 or longer.

[0141] <Flowchart of the abnormality detection process> The abnormality detection process in Embodiment 2 will now be described. Figure 18 is a flowchart showing the abnormality detection process according to Embodiment 2. First, the difference WDiff** between each output pulse width included in the output pulse width Wact** and each command pulse width included in the command pulse width Wref** is calculated, and the respective difference WDiff** is compared with the threshold Dth (Step ST2061: Pulse width comparison process).

[0142] Next, for each difference included in the difference WDiff**, if it is greater than or equal to the threshold Dth, it is determined that there is an anomaly, and if the difference WDiff is less than the threshold Dth, it is determined that there is no anomaly, thereby obtaining the difference anomaly determination result ErrD** as the determination result (Step ST2062: Difference Anomaly Determination Step).

[0143] Next, for each output pulse width included in the output pulse width Wact**, the time during which the same value persists is measured as the pulse width fixation time FixT** (Step ST2063: Pulse width fixation time measurement process). If the PWM pulse period is varied, the upper limit of the output pulse width changes, so the time during which the upper limit of the output pulse width persists is also measured as the fixation time.

[0144] Next, based on the torque command Trq* output by the higher-level controller or system (not shown in the diagram), it is determined whether there is a time variation in the ideal value of the PWM control command (three-phase duty cycle commands Dref_u, Dref_v, Dref_w), and the presence or absence of time variation Ref_v is obtained as the determination result (Step ST2064: Rotational variation determination step). In the rotational variation determination step, if the torque command Trq* is zero, it is determined that there is "no time variation", and if the torque command Trq* is not zero, it is determined that there is "time variation".

[0145] Next, based on the pulse width fixing time FixT**, the electrical angle period Tω, and the presence or absence of time variation Ref_v, the presence or absence of a control command abnormality is determined, and the control command abnormality determination result ErrR** is obtained as the determination result (Step ST2065: Control command abnormality determination step). In the control command abnormality determination step, if the presence or absence of time variation Ref_v is "time variation present" and the pulse width fixing time FixT** is Tω / 2 or greater, it is determined that "there is a control command abnormality". If the above conditions are not met, that is, if the presence or absence of time variation Ref_v is "no time variation" or the pulse width fixing time FixT** is less than Tω / 2, it is determined that "there is no control command abnormality". As mentioned above, the control command abnormality determination step may not be performed when the rotational speed of the motor 901 is less than or equal to a predetermined rotational speed. In this case, the control command abnormality determination step is performed only when the rotational speed of the motor 901 is greater than the predetermined rotational speed.

[0146] Next, the presence or absence of an abnormality is determined based on the differential abnormality determination result ErrD** and the control command abnormality determination result ErrR**, and the abnormality determination result Err** is obtained as the determination result (step ST2066: abnormality presence / absence determination step). In the abnormality presence / absence determination step, if at least one of the differential abnormality determination result ErrD** and the control command abnormality determination result ErrR** is found to be abnormal, it is determined that there is an abnormality, and if neither is found to be abnormal, it is determined that there is no abnormality. As described above, since both the differential abnormality determination result ErrD** and the control command abnormality determination result ErrR** contain six determination results corresponding to each switching element of the inverter circuit 902, the presence or absence of an abnormality is also determined for each of them.

[0147] Next, based on the output pulse width Wact** and pulse width lock time FixT**, the presence or absence of ON lock for each switching element of the inverter circuit 902 is determined, and the ON lock determination result FixON** is obtained as the determination result. (Step ST2067: ON lock determination step). In the ON lock determination step, if the pulse width lock time FixT** is greater than or equal to a predetermined time (e.g., Tω / 2 or more) and the output pulse width Wact** is the PWM pulse period, it is determined that ON lock has occurred.

[0148] Next, based on the output pulse width Wact** and pulse width lock time FixT**, the presence or absence of off-locking for each switching element of the inverter circuit 902 is determined, and the off-locking determination result FixOFF** is obtained as the determination result (Step ST2068: Off-locking determination step). In the off-locking determination step, if the pulse width lock time FixT** is greater than or equal to a predetermined time (for example, Tω / 2 or more) and the output pulse width Wact** is zero, it is determined that off-locking has occurred.

[0149] Next, the state of the PWM control is determined based on the abnormality determination result Err**, the on-fixed determination result FixON**, and the off-fixed determination result FixOFF**, and the state determination result ErrSt** is obtained as the determination result (Step ST2069: Control state determination step). As described above, the state determination result ErrSt** has the following possibilities: "Normal", "On-fixed abnormality", "Off-fixed abnormality", and "Pulse abnormality".

[0150] According to Embodiment 2, the same effects as in Embodiment 1 can be obtained in a PWM control device applied to a power conversion device that controls an electric motor. Furthermore, the system is configured to detect abnormalities in the PWM control command (three-phase duty cycle command) based on the motor's rotational state and the ideal value of the PWM control command. Therefore, even if there is an abnormality in the command pulse width and it is not possible to correctly determine the abnormality of the PWM control by comparing only the output pulse width and the command pulse width, the system can still detect the abnormality in the PWM control command and correctly determine the abnormality.

[0151] Furthermore, the system performs an abnormality check on each of the multiple PWM control signals, and if an abnormality is detected and the system instructs the output of the PWM control signal to be stopped, it sets the signal level at the time of output stop according to the type of abnormality and the PWM control signal in which the abnormality occurred. This allows the motor to be stopped appropriately.

[0152] Embodiment 3. Next, Embodiment 3 will be described with reference to Figures 19 to 22. Embodiment 3 is the same as Embodiment 2 in terms of the configuration of the power converter and the overall configuration of the PWM control device, but the abnormality determination means omits the implementation of differential abnormality determination, and the configuration of the control unit and abnormality determination means differs from Embodiment 2.

[0153] The configuration and operation of the PWM control device of Embodiment 3 will be described below, focusing on the differences from Embodiment 2. Parts identical or equivalent to those in Embodiment 2 are denoted by the same reference numerals.

[0154] <Functional block of the control unit> Figure 19 is a block diagram showing the configuration of the control unit according to Embodiment 3. The control unit 30 omits the command pulse width calculation means 23 from the control unit 20 of Embodiment 2. Therefore, there is no input of the command pulse width Wref** to the abnormality determination means 37.

[0155] The abnormality determination means 37 will now be described. Figure 20 is a block diagram showing the configuration of the abnormality determination means according to Embodiment 3. The abnormality determination means 37 is the same as the abnormality determination means 27 of Embodiment 2, but with the pulse width comparison unit 27a and the difference abnormality determination unit 27b omitted. The pulse width fixing time measurement unit 27c, rotation fluctuation determination unit 27d, and control command abnormality determination unit 27e are the same as in Embodiment 2. The abnormality presence / absence determination unit 37f determines the presence or absence of an abnormality based only on the control command abnormality determination result ErrR**, and outputs the result as the abnormality determination result Err** to the control state determination unit 27i. The on-fixation determination unit 27g, off-fixation determination unit 27h, and control state determination unit 27i are the same as in Embodiment 2.

[0156] <Operation of the PWM control device> Next, the operation will be described. FIG. 21 is a flowchart showing the operation of the PWM control device in Embodiment 3. As can be seen by comparing FIG. 17 and FIG. 21, in the operation of the PWM control device in Embodiment 3, the command pulse width calculation step is omitted. Steps ST301 to ST304 correspond to steps ST201 to ST204, and steps ST305 to ST308 correspond to steps ST206 to ST209, and each step is the same as in Embodiment 2.

[0157] <Flowchart of the abnormality determination step> The abnormality determination step in Embodiment 3 will be described. FIG. 22 is a flowchart showing the abnormality determination step according to Embodiment 3. As can be seen by comparing FIG. 18 and FIG. 22, in the abnormality determination step in Embodiment 3, only the pulse width comparison step and the differential abnormality determination step are omitted, and steps ST3051 to ST3057 correspond to steps ST2063 to ST2069. For each step, the difference from step ST2066 is that in the abnormality presence / absence determination step of step ST3054, the presence / absence of an abnormality is determined based only on the control command abnormality determination result ErrR**, but the other steps are the same as in Embodiment 2.

[0158] According to Embodiment 3, by not performing the determination of differential abnormality, the calculation of the command pulse width is also omitted. Therefore, while simplifying the configuration compared to Embodiment 2, it is possible to detect an abnormality in PWM control by detecting an abnormality in the PWM control command.

[0159] In addition, in the third embodiment, the abnormal determination of the control command, the on-sticking determination, and the off-sticking determination are made based on the output pulse width Wact**. However, it is also conceivable to use the command pulse width Wref** instead of the output pulse width Wact**. In this case, it is necessary to add command pulse width calculation means to the control unit, but the output pulse width acquisition means can be omitted. That is, in the third embodiment, at least one of the output pulse width acquisition means 22 and the command pulse width calculation means 23 may be provided. Regarding the abnormal determination means 37, when using the command pulse width Wref**, the output pulse width Wact** in FIG. 20 may be replaced with the command pulse width Wref**. The same applies to the abnormal determination process.

[0160] Fourth Embodiment. Next, the fourth embodiment will be described based on FIGS. 23 and 24. The fourth embodiment is obtained by omitting the on-sticking determination unit and the off-sticking determination unit from the abnormal determination means of the third embodiment. FIG. 23 is a block diagram showing the configuration of the abnormal determination means according to the fourth embodiment. The abnormal determination means 47 omits the on-sticking determination unit 27g and the off-sticking determination unit 27h from the abnormal determination means 37 of the third embodiment. In addition, with the omission of the on-sticking determination unit 27g and the off-sticking determination unit 27h, the on-sticking determination result FixON** and the off-sticking determination result FixOFF** are also eliminated. Therefore, the state determination result ErrSt** has only "abnormal" or "normal", and is substantially the same as the abnormal determination result Err**. For this reason, the control state determination unit 27i is also omitted, and the presence / absence of abnormality determination unit 47f is configured to output the state determination result ErrSt**.

[0161] <Operation of PWM Control Device (Flow of Abnormal Determination Process)> Next, the operation will be described. Except for the abnormality determination process, it is the same as in Embodiment 3, so only the flow of the abnormality determination process will be described. Figure 24 is a flowchart showing the abnormality determination process according to Embodiment 4. As can be seen by comparing Figure 22 and Figure 24, in the abnormality determination process in Embodiment 4, only the ON-fixation determination process, OFF-fixation determination process, and control state determination process are omitted, and steps ST4051 to ST4054 correspond to steps ST3051 to ST3054. However, as mentioned above, in Embodiment 4, the state determination result ErrSt** is substantially the same as Err**, so the abnormality determination process in step ST4054 also includes the control state determination in step ST3057. The other processes are the same as in Embodiment 3.

[0162] According to Embodiment 4, by omitting both the on-fixation determination and the off-fixation determination, the configuration is made even simpler than in Embodiment 3, while still enabling the detection of abnormalities in PWM control by detecting abnormalities in the PWM control command.

[0163] In Embodiment 4, as in Embodiment 3, it is possible to use the command pulse width Wref** instead of the output pulse width Wact**.

[0164] Although this application describes various exemplary embodiments and examples, the various features, aspects, and functions described in one or more embodiments are not limited to the application of a particular embodiment, but can be applied individually or in various combinations to the embodiments. Accordingly, countless variations not illustrated are conceivable within the scope of the art disclosed herein. These include, for example, modifying, adding or omitting at least one component, or even extracting at least one component and combining it with components of other embodiments.

[0165] For example, while Embodiment 4 is configured to only determine abnormalities in control commands, it is also conceivable that abnormalities in PWM control can be detected by determining only ON-fixation or only OFF-fixation.

[0166] Furthermore, it is also conceivable to apply a normal control determination means, such as that in Embodiment 2, to Embodiment 1. In this case, a threshold value for normal determination is set separately, and if the state determination result, which is the result of the abnormal determination, remains "normal" for a period of time equal to or longer than this threshold value, it is determined that the PWM control is operating normally.

[0167] Furthermore, the predetermined time for determining whether there is "on-locking" or "off-locking" in the on-locking determination means and off-locking determination means shown in Embodiments 2 and 3 does not necessarily have to be based on a value indicating the rotation state of the electric motor. For this reason, on-locking determination and off-locking determination can be performed even if the controlled device does not have an electric motor. For example, it is conceivable to add a pulse width locking time measurement unit, an on-locking determination unit, and an off-locking determination unit to the abnormality determination means of Embodiment 1. In this case, if an abnormality in PWM control is determined from the determination result of the differential abnormality determination, it is determined whether the abnormality is an "on-locking abnormality" or an "off-locking abnormality". As mentioned above, if there is an abnormality in PWM control but neither on-locking nor off-locking occurs, it will be classified as a "pulse abnormality".

[0168] The various aspects of this disclosure are summarized below as an appendix. (Note 1) A control command generation means for generating PWM control commands, A control signal generation means that generates a PWM control signal for controlling a controlled device based on the PWM control command, An output pulse width acquisition means for acquiring the output pulse width of the PWM control signal, A command pulse width calculation means for calculating the command pulse width, which is the pulse width of the PWM control command, The system includes an abnormality determination means for determining whether or not there is an abnormality in PWM control by comparing the output pulse width and the command pulse width, and the output pulse width acquisition means is A PWM control device characterized by measuring an output pulse width measurement value, which is a measured value of the width of the output pulse of the PWM control signal, based on the edge of the PWM control signal, and determining whether or not there is an output pulse of the PWM control signal. If it is determined that there is no output pulse, the output pulse width measurement value is corrected, and the corrected output pulse width measurement value is set as the output pulse width. (Note 2) The output pulse width acquisition means is An output pulse edge detection unit that detects the aforementioned edge, An output pulse width measuring unit that measures the output pulse width measurement value based on the edge, A pulse presence / absence determination unit that determines whether or not there is an output pulse of the PWM control signal, The PWM control device according to Appendix 1, further comprising an output pulse width correction unit that corrects the output pulse width measurement value when it is determined that there is no output pulse, and sets the corrected output pulse width measurement value as the output pulse width. (Note 3) The output pulse width acquisition means further comprises an edge time acquisition unit that acquires the detection time of the edge, The pulse presence / absence determination unit determines the presence or absence of the output pulse based on the difference between the last edge detection time, which is the time when the edge was last detected, and the current time, as described in Appendix 2 of the PWM control device. (Note 4) The output pulse width acquisition means further comprises a signal level acquisition unit that acquires the signal level of the PWM control signal, The PWM control device according to Appendix 3, wherein the output pulse width correction unit corrects the output pulse width measurement value based on the signal level of the PWM control signal at the final edge detection time. (Note 5) The control signal generation means generates the PWM control signal based on the PWM pulse period, The PWM control device according to Appendix 4, wherein the output pulse width correction unit corrects the output pulse width measurement value to the PWM pulse period when the signal level of the PWM control signal at the final edge detection time is "high", and corrects the output pulse width measurement value to zero when the signal level of the PWM control signal at the final edge detection time is "low". (Note 6) The control signal generation means generates the PWM control signal based on the PWM pulse period, The PWM control device according to Appendix 3 or 4, wherein the pulse presence / absence determination unit determines that there is no output pulse when the difference between the current time and the final edge detection time is equal to or greater than the PWM pulse period. (Note 7) The controlled device has an electric motor that operates by being supplied with power in accordance with the PWM control signal, The PWM control device further comprises a rotation state acquisition means for acquiring the rotation state of the electric motor, The abnormality determination means includes a pulse width fixing time measuring unit that measures a pulse width fixing time which is the time during which the output pulse width or the command pulse width remains constant, and determines whether or not there is an abnormality in the PWM control based on a comparison of a time determined based on the rotation state and the pulse width fixing time, and whether or not there is a time variation in the ideal value of the PWM control command, as described in any one of the appendices 1 to 6. (Note 8) The rotation state acquisition means acquires at least the electrical angular period of the electric motor, The PWM control device according to Appendix 7, wherein the abnormality determination means determines that there is an abnormality in the PWM control when there is a time variation in the ideal value and the pulse width fixing time is 1 / 2 or more of the electrical angular period. (Note 9) The system further includes a pulse width fixing time measuring unit that measures the pulse width fixing time, which is the time during which the output pulse width or the command pulse width remains constant. If the abnormality determination means determines that there is an abnormality in the PWM control, and the pulse width fixing time is longer than or equal to a predetermined time, A PWM control device according to any one of the appendices 1 to 6, wherein if the constant value is the PWM pulse period, the abnormality of the PWM control is determined to be an on-fixation abnormality, and if the constant value is zero, the abnormality of the PWM control is determined to be an off-fixation abnormality. (Note 10) The controlled device has an electric motor that operates by being supplied with power in accordance with the PWM control signal, The PWM control device is A rotation state acquisition means for acquiring the rotation state of the electric motor, The system further includes a normal control determination means for determining whether the PWM control by the PWM control device is operating normally, The PWM control device according to any one of the appendices 1 to 6, wherein the normal control determination means determines that the PWM control is operating normally if the state in which the determination result of the abnormality determination means is normal continues for a period of time or longer based on the rotation state. (Note 11) The PWM control device according to Appendix 10, wherein the rotation state acquisition means acquires at least the electrical angular period of the electric motor, and the time determined based on the rotation state is half of the electrical angular period. (Note 12) The PWM control device according to any one of the appendices 1 to 11, further comprising an output stop instruction means for instructing the control signal generation means to stop outputting the PWM control signal when the abnormality determination means determines that there is an abnormality in the PWM control by the PWM control device. (Note 13) The control signal generation means generates a plurality of PWM control signals, The PWM control device according to Appendix 12, wherein the output stop instruction means sets the signal level of each PWM control signal when stopping the output of each PWM control signal in accordance with an abnormality of each PWM control signal. (Note 14) A PWM control device that outputs the PWM control signal to a controlled device having an electric motor that operates by being supplied with power in accordance with the PWM control signal, A control command generation means for generating PWM control commands, A control signal generation means that generates the PWM control signal based on the PWM control command, At least one of the following: an output pulse width acquisition means for acquiring the output pulse width of the PWM control signal, and a command pulse width calculation means for calculating the command pulse width, which is the pulse width of the PWM control command; An abnormality detection means for determining whether or not there is an abnormality in PWM control, The system includes a rotation state acquisition means for acquiring the rotation state of the electric motor, The abnormality determination means includes a pulse width fixing time measuring unit that measures pulse width fixing time, which is the time during which the output pulse width or the command pulse width remains constant. The PWM control device is characterized in that it determines whether or not there is an abnormality in the PWM control based on a comparison of a time determined based on the rotation state and the pulse width fixing time, and whether or not there is a time variation in the ideal value of the PWM control command. (Note 15) If the abnormality determination means determines that there is an abnormality in the PWM control, and the pulse width fixing time is longer than or equal to a predetermined time, The PWM control device described in Appendix 14, wherein if the constant value is the PWM pulse period, the abnormality of the PWM control is determined to be an on-fixation abnormality, and if the constant value is zero, the abnormality of the PWM control is determined to be an off-fixation abnormality. (Note 16) A control command generation process that generates PWM control commands, A control signal generation step that generates a PWM control signal for controlling a controlled device based on the PWM control command, An output pulse width acquisition step for acquiring the output pulse width of the PWM control signal, A command pulse width calculation step for calculating the command pulse width, which is the pulse width of the PWM control command, The system includes an abnormality determination step that determines whether or not there is an abnormality in PWM control by comparing the output pulse width and the command pulse width, and in the output pulse width acquisition step, A PWM control method characterized by measuring an output pulse width measurement value, which is a measured value of the width of the output pulse of the PWM control signal, based on the edge of the PWM control signal, and determining whether or not there is an output pulse of the PWM control signal. If it is determined that there is no output pulse, the output pulse width measurement value is corrected, and the corrected output pulse width measurement value is set as the output pulse width. (Note 17) The output pulse width acquisition step includes an output pulse edge detection step for detecting the edge, An output pulse width measurement step of measuring the output pulse width measurement value based on the edge, A step to determine whether or not there is an output pulse of the PWM control signal, The PWM control method according to Appendix 16, which includes an output pulse width correction step in which, if it is determined that there is no output pulse, the output pulse width measurement value is corrected and the corrected output pulse width measurement value is set as the output pulse width. (Note 18) The output pulse width acquisition step further includes an edge time acquisition step for acquiring the detection time of the edge, The PWM control method according to Appendix 17, wherein in the output pulse presence / absence determination step, the presence or absence of the output pulse is determined based on the difference between the last edge detection time, which is the time when the edge was last detected, and the current time. (Note 19) The output pulse width acquisition step further includes a signal level acquisition step for acquiring the signal level of the PWM control signal, The PWM control method according to Appendix 18, wherein in the output pulse width correction step, the output pulse width measurement value is corrected based on the signal level of the PWM control signal at the final edge detection time. (Note 20) In the control signal generation step, the PWM control signal is generated by the PWM pulse period, The PWM control method according to Appendix 19, wherein in the output pulse width correction step, if the signal level of the PWM control signal at the final edge detection time is "high", the output pulse width measurement value is corrected to the PWM pulse period, and if the signal level of the PWM control signal at the final edge detection time is "low", the output pulse width measurement value is corrected to zero. (Note 21) In the control signal generation step, the PWM control signal is generated by the PWM pulse period, The PWM control method according to Appendix 18 or 19, wherein in the output pulse presence / absence determination step, if the difference between the current time and the final edge detection time is greater than or equal to the PWM pulse period, it is determined that there is no output pulse. (Note 22) The controlled device has an electric motor that operates by being supplied with power in accordance with the PWM control signal, The PWM control method further includes a rotation state acquisition step for acquiring the rotation state of the electric motor, The PWM control method according to any one of the appendices 16 to 21, wherein in the abnormality determination step, the pulse width fixing time is the time during which the output pulse width or the command pulse width remains constant, and the presence or absence of an abnormality in the PWM control is determined by comparing the pulse width fixing time with a time determined based on the rotation state, and by determining whether or not there is a time variation in the ideal value of the PWM control command. (Note 23) In the rotation state acquisition step, at least the electrical angular period of the electric motor is acquired, The PWM control method according to Appendix 22, wherein in the abnormality determination step, if there is a time variation in the ideal value and the pulse width fixing time is 1 / 2 or more of the electrical angular period, it is determined that there is an abnormality in the PWM control. (Note 24) The process further includes measuring the pulse width fixation time, which is the time during which the output pulse width or the command pulse width remains constant. If, in the abnormality determination step, it is determined that there is an abnormality in the PWM control, and the pulse width fixing time is longer than or equal to a predetermined time, The PWM control method according to any one of the appendices 16 to 21, wherein if the constant value is the PWM pulse period, the abnormality of the PWM control is determined to be an on-fixation abnormality, and if the constant value is zero, the abnormality of the PWM control is determined to be an off-fixation abnormality. (Note 25) The controlled device has an electric motor that operates by being supplied with power in accordance with the PWM control signal, The PWM control method is as follows: A rotation state acquisition step for acquiring the rotation state of the electric motor, The system further includes a normal control determination step for determining whether the PWM control by the PWM control method described above is operating normally, The PWM control method according to any one of the appendices 16 to 21, wherein in the normal control determination step, if the state of no abnormality as determined in the abnormality determination step continues for a period of time or longer based on the rotation state, it is determined that the PWM control is operating normally. (Note 26) The PWM control method according to Appendix 25, wherein in the rotation state acquisition step, at least the electrical angular period of the electric motor is acquired, and the time determined based on the rotation state is half of the electrical angular period. (Note 27) The PWM control method according to any one of the appendices 16 to 26, wherein, in the abnormality determination step, if it is determined that there is an abnormality in the PWM control by the PWM control method, the output of the PWM control signal is stopped. (Note 28) In the control signal generation step, a plurality of PWM control signals are generated. The PWM control method described in Appendix 27, wherein when the output of the PWM control signal is stopped, the signal level of each PWM control signal is set according to the abnormality of each PWM control signal. (Note 29) A PWM control method that outputs the PWM control signal to a control device having an electric motor that operates by being supplied with power in accordance with the PWM control signal, A control command generation process that generates PWM control commands, A control signal generation step that generates the PWM control signal based on the PWM control command, At least one of the following steps: an output pulse width acquisition step for acquiring the output pulse width of the PWM control signal, and a command pulse width calculation step for calculating the command pulse width, which is the pulse width of the PWM control command. An abnormality detection step to determine whether or not there is an abnormality in PWM control, The system includes a rotation state acquisition step for acquiring the rotation state of the electric motor, A PWM control method characterized in that, in the abnormality determination step, the pulse width fixation time is the time during which the output pulse width or the command pulse width remains constant, and the presence or absence of an abnormality in the PWM control is determined by comparing the pulse width fixation time with a time determined based on the rotation state, and by the presence or absence of time fluctuations in the ideal value of the PWM control command. (Note 30) If, in the abnormality determination step, it is determined that there is an abnormality in the PWM control, and the pulse width fixing time is longer than or equal to a predetermined time, The PWM control method described in Appendix 29, wherein if the constant value is the PWM pulse period, the abnormality of the PWM control is determined to be an on-fixation abnormality, and if the constant value is zero, the abnormality of the PWM control is determined to be an off-fixation abnormality. [Explanation of Symbols]

[0169] 10, 20, 30 Control unit; 11, 21 Control command generation means; 11a, 21e Duty command calculation unit; 12, 22 Output pulse width acquisition means; 12a Output pulse edge detection unit; 12b Output pulse width measurement unit; 12c Edge time acquisition unit; 12d Pulse presence / absence determination unit; 12e Signal level acquisition unit; 12f Output pulse width correction unit; 13, 23 Command pulse width calculation means; 17, 27, 37, 47 Anomaly determination means; 17a, 27a Pulse width comparison unit; 17b, 27b Difference anomaly determination unit; 18, 28 Output stop instruction means; 19, 29 Control signal generation means; 25 Rotation state acquisition means; 25b Electrical angle period calculation unit; 27c Pulse width fixation time measurement unit; 27e Control command anomaly determination unit; 27f, 37f, 47f Anomaly presence / absence determination unit; 27g ON-lock determination unit, 27h OFF-lock determination unit, 100, 200 PWM control device, 900 Controlled device, 901 Motor, Dref Duty command, Dref_u, Dref_v, Dref_w Three-phase duty command, EdgLev Signal level, EdgR Pulse presence / absence determination result, EdgT Edge time, ErrSt, ErrSt** State determination result, FixT** Pulse width lock time, Ft Last edge detection time, Pt Current time, Stp Stop instruction signal, SW, SW_UH, SW_VH, SW_WH, SW_UL, SW_VL, SW_WL, SW** PWM control signal, Tω Electrical angle period, Wact, Wact** Output pulse width, Wref, Wref** Command pulse width, Wmon Output pulse width measurement value

Claims

1. A control command generation means for generating PWM control commands, A control signal generation means that generates a PWM control signal to control a controlled device based on the PWM control command, An output pulse width acquisition means for acquiring the output pulse width of the PWM control signal, A command pulse width calculation means for calculating the command pulse width, which is the pulse width of the PWM control command, The system includes an abnormality determination means that determines whether or not there is an abnormality in PWM control by comparing the output pulse width and the command pulse width, and the output pulse width acquisition means is A PWM control device characterized by measuring an output pulse width measurement value, which is a measured value of the width of the output pulse of the PWM control signal, based on the edge of the PWM control signal, and determining whether or not there is an output pulse of the PWM control signal. If it is determined that there is no output pulse, the output pulse width measurement value is corrected, and the corrected output pulse width measurement value is set as the output pulse width.

2. The output pulse width acquisition means is An output pulse edge detection unit that detects the aforementioned edge, An output pulse width measuring unit that measures the output pulse width measurement value based on the edge, A pulse presence / absence determination unit that determines the presence or absence of the output pulse of the PWM control signal, The PWM control device according to claim 1, further comprising: an output pulse width correction unit that corrects the output pulse width measurement value when it is determined that there is no output pulse, and sets the corrected output pulse width measurement value as the output pulse width.

3. The output pulse width acquisition means further comprises an edge time acquisition unit that acquires the detection time of the edge, The PWM control device according to claim 2, wherein the pulse presence / absence determination unit determines the presence or absence of the output pulse based on the difference between the last edge detection time, which is the time when the edge was last detected, and the current time.

4. The output pulse width acquisition means further comprises a signal level acquisition unit that acquires the signal level of the PWM control signal, The PWM control device according to claim 3, wherein the output pulse width correction unit corrects the output pulse width measurement value based on the signal level of the PWM control signal at the final edge detection time.

5. The control signal generation means generates the PWM control signal based on the PWM pulse period, The PWM control device according to claim 4, wherein the output pulse width correction unit corrects the output pulse width measurement value to the PWM pulse period when the signal level of the PWM control signal at the final edge detection time is "high", and corrects the output pulse width measurement value to zero when the signal level of the PWM control signal at the final edge detection time is "low".

6. The control signal generation means generates the PWM control signal based on the PWM pulse period, The PWM control device according to claim 3 or 4, wherein the pulse presence / absence determination unit determines that there is no output pulse when the difference between the current time and the final edge detection time is equal to or greater than the PWM pulse period.

7. The controlled device has an electric motor that operates by being supplied with power in accordance with the PWM control signal, The PWM control device further comprises a rotation state acquisition means for acquiring the rotation state of the electric motor, The abnormality determination means includes a pulse width fixing time measuring unit that measures a pulse width fixing time which is the time during which the output pulse width or the command pulse width remains constant, and determines whether or not there is an abnormality in the PWM control based on a comparison of a time determined based on the rotation state and the pulse width fixing time, and whether or not there is a time variation in the ideal value of the PWM control command, as described in any one of claims 1 to 5.

8. The rotation state acquisition means acquires at least the electrical angular period of the electric motor, The PWM control device according to claim 7, wherein the abnormality determination means determines that there is an abnormality in the PWM control when there is a time variation in the ideal value and the pulse width fixing time is half or more of the electrical angular period.

9. The system further includes a pulse width fixing time measuring unit that measures the pulse width fixing time, which is the time during which the output pulse width or the command pulse width remains constant. If the abnormality determination means determines that there is an abnormality in the PWM control, and the pulse width fixing time is longer than or equal to a predetermined time, A PWM control device according to any one of claims 1 to 5, wherein if the constant value is the PWM pulse period, the abnormality of the PWM control is determined to be an on-lock abnormality, and if the constant value is zero, the abnormality of the PWM control is determined to be an off-lock abnormality.

10. The controlled device has an electric motor that operates by being supplied with power in accordance with the PWM control signal, The PWM control device is A rotation state acquisition means for acquiring the rotation state of the electric motor, The system further comprises a normal control determination means for determining whether or not the PWM control by the PWM control device is operating normally, The PWM control device according to any one of claims 1 to 5, wherein the normal control determination means determines that the PWM control is operating normally if the state in which the determination result of the abnormality determination means is normal continues for a period of time or longer based on the rotation state.

11. The PWM control device according to claim 10, wherein the rotation state acquisition means acquires at least the electrical angular period of the electric motor, and the time determined based on the rotation state is half of the electrical angular period.

12. The PWM control device according to any one of claims 1 to 5, further comprising an output stop instruction means for instructing the control signal generation means to stop outputting the PWM control signal when the abnormality determination means determines that there is an abnormality in the PWM control by the PWM control device.

13. The control signal generation means generates a plurality of PWM control signals, The PWM control device according to claim 12, wherein the output stop instruction means sets the signal level of each PWM control signal when stopping the output of each PWM control signal in response to an abnormality in each of the PWM control signals.

14. A PWM control device that outputs the PWM control signal to a controlled device having an electric motor that operates by being supplied with power according to the PWM control signal, A control command generation means for generating PWM control commands, A control signal generation means that generates the PWM control signal based on the PWM control command, At least one of the following: an output pulse width acquisition means for acquiring the output pulse width of the PWM control signal, and a command pulse width calculation means for calculating the command pulse width, which is the pulse width of the PWM control command; An abnormality detection means for determining whether or not there is an abnormality in PWM control, The system includes a rotation state acquisition means for acquiring the rotation state of the electric motor, The abnormality determination means includes a pulse width fixing time measuring unit that measures pulse width fixing time, which is the time during which the output pulse width or the command pulse width remains constant. The PWM control device is characterized in that it determines whether or not there is an abnormality in the PWM control based on a comparison of a time determined based on the rotation state and the pulse width fixing time, and whether or not there is a time fluctuation in the ideal value of the PWM control command.

15. If the abnormality determination means determines that there is an abnormality in the PWM control, and the pulse width fixing time is longer than or equal to a predetermined time, The PWM control device according to claim 14, wherein if the constant value is the PWM pulse period, the abnormality of the PWM control is determined to be an on-fixed abnormality, and if the constant value is zero, the abnormality of the PWM control is determined to be an off-fixed abnormality.

16. A control command generation process that generates PWM control commands, A control signal generation step that generates a PWM control signal to control a controlled device based on the PWM control command, An output pulse width acquisition step for acquiring the output pulse width of the PWM control signal, A command pulse width calculation step for calculating the command pulse width, which is the pulse width of the PWM control command, The system includes an abnormality determination step which determines whether or not there is an abnormality in PWM control by comparing the output pulse width and the command pulse width, and in the output pulse width acquisition step, A PWM control method characterized by measuring an output pulse width measurement value, which is a measured value of the width of the output pulse of the PWM control signal, based on the edge of the PWM control signal, and determining whether or not there is an output pulse of the PWM control signal. If it is determined that there is no output pulse, the output pulse width measurement value is corrected, and the corrected output pulse width measurement value is set as the output pulse width.

17. The output pulse width acquisition step includes an output pulse edge detection step for detecting the edge, An output pulse width measurement step of measuring the output pulse width measurement value based on the edge, An output pulse presence / absence determination step for determining the presence or absence of an output pulse of the PWM control signal, The PWM control method according to claim 16, which includes an output pulse width correction step in which, if it is determined that there is no output pulse, the output pulse width measurement value is corrected and the corrected output pulse width measurement value is set as the output pulse width.

18. The output pulse width acquisition step further includes an edge time acquisition step for acquiring the detection time of the edge, The PWM control method according to claim 17, wherein in the output pulse presence / absence determination step, the presence or absence of the output pulse is determined based on the difference between the last edge detection time, which is the time when the edge was last detected, and the current time.

19. The output pulse width acquisition step further includes a signal level acquisition step for acquiring the signal level of the PWM control signal, The PWM control method according to claim 18, wherein in the output pulse width correction step, the output pulse width measurement value is corrected based on the signal level of the PWM control signal at the final edge detection time.

20. In the control signal generation step, the PWM control signal is generated by the PWM pulse period. The PWM control method according to claim 19, wherein in the output pulse width correction step, if the signal level of the PWM control signal at the final edge detection time is "high", the output pulse width measurement value is corrected to the PWM pulse period, and if the signal level of the PWM control signal at the final edge detection time is "low", the output pulse width measurement value is corrected to zero.

21. In the control signal generation step, the PWM control signal is generated by the PWM pulse period. The PWM control method according to claim 18 or 19, wherein in the output pulse presence / absence determination step, if the difference between the current time and the final edge detection time is greater than or equal to the PWM pulse period, it is determined that there is no output pulse.

22. The controlled device has an electric motor that operates by being supplied with power in accordance with the PWM control signal, The PWM control method further includes a rotation state acquisition step of acquiring the rotation state of the electric motor, The PWM control method according to any one of claims 16 to 20, wherein in the abnormality determination step, the pulse width fixing time is the time during which the output pulse width or the command pulse width remains constant, and the presence or absence of an abnormality in the PWM control is determined based on a comparison of the time determined based on the rotation state and the pulse width fixing time, and the presence or absence of time fluctuation of the ideal value of the PWM control command.

23. In the rotation state acquisition step, at least the electrical angular period of the electric motor is acquired, The PWM control method according to claim 22, wherein in the abnormality determination step, if there is a time variation in the ideal value and the pulse width fixing time is half or more of the electrical angle period, it is determined that there is an abnormality in the PWM control.

24. The process further includes measuring the pulse width fixation time, which is the time during which the output pulse width or the command pulse width remains constant. If, in the abnormality determination step, it is determined that there is an abnormality in the PWM control, and the pulse width fixing time is longer than or equal to a predetermined time, The PWM control method according to any one of claims 16 to 20, wherein if the constant value is the PWM pulse period, the abnormality of the PWM control is determined to be an on-lock abnormality, and if the constant value is zero, the abnormality of the PWM control is determined to be an off-lock abnormality.

25. The controlled device has an electric motor that operates by being supplied with power in accordance with the PWM control signal, The PWM control method is as follows: A rotation state acquisition step for acquiring the rotation state of the electric motor, The system further comprises a normal control determination step for determining whether the PWM control by the PWM control method described above is operating normally, The PWM control method according to any one of claims 16 to 20, wherein in the normal control determination step, if the state of no abnormality as determined in the abnormality determination step continues for a period of time or longer based on the rotation state, it is determined that the PWM control is operating normally.

26. The PWM control method according to claim 25, wherein in the rotation state acquisition step, at least the electrical angular period of the electric motor is acquired, and the time determined based on the rotation state is half of the electrical angular period.

27. The PWM control method according to any one of claims 16 to 20, wherein, in the abnormality determination step, if it is determined that there is an abnormality in the PWM control by the PWM control method, the output of the PWM control signal is stopped.

28. In the control signal generation step, a plurality of PWM control signals are generated. The PWM control method according to claim 27, wherein when the output of the PWM control signal is stopped, the signal level of each PWM control signal is set according to the abnormality of each PWM control signal.

29. A PWM control method that outputs the PWM control signal to a control device having an electric motor that operates by being supplied with power according to the PWM control signal, A control command generation process that generates PWM control commands, A control signal generation step that generates the PWM control signal based on the PWM control command, At least one of the following steps: an output pulse width acquisition step for acquiring the output pulse width of the PWM control signal, and a command pulse width calculation step for calculating the command pulse width, which is the pulse width of the PWM control command. An abnormality detection process for determining whether or not there is an abnormality in PWM control, The system includes a rotation state acquisition step for acquiring the rotation state of the electric motor, A PWM control method characterized in that, in the abnormality determination step, the pulse width fixing time is the time during which the output pulse width or the command pulse width remains constant, and the presence or absence of an abnormality in the PWM control is determined based on a comparison of the time determined based on the rotation state and the pulse width fixing time, and the presence or absence of time fluctuation of the ideal value of the PWM control command.

30. If, in the abnormality determination step, it is determined that there is an abnormality in the PWM control, and the pulse width fixing time is longer than or equal to a predetermined time, The PWM control method according to claim 29, wherein if the constant value is the PWM pulse period, the abnormality of the PWM control is determined to be an on-lock abnormality, and if the constant value is zero, the abnormality of the PWM control is determined to be an off-lock abnormality.

Images