Semiconductor device

Abstract

To provide a semiconductor device capable of preventing the malfunction of a system mounted with a tracking filter.SOLUTION: The tracking filter TF includes multipliers ML1x and ML1y for extraction, low pass filters LPF Fx and LPFy, and multipliers ML2x and ML2y for restoration. The multiplier for extraction multiplies the tracking input signal Fin by a cosine wave signal or a sine wave signal of the tracking frequency (ω r). The multiplier for restoration generates a tracking signal Fout by multiplying the signals F2x and F2y outputted from the low pass filter by a cosine-wave signal or a sine-wave signal of the tracking frequency (ω r). A monitoring circuit MNI detects AC components ACx included in the signals F2x and F2y outputted from the filter TF and judges whether or not the filter TF is abnormal based on the magnitude of the components ACx.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a semiconductor device, for example, a semiconductor device having a tracking filter. [Background technology]

[0002] Patent Document 1 discloses a control device that can suppress vibrations at frequencies proportional to the rotational speed of a motor. The control device detects vibrations using an acceleration sensor installed at a predetermined location and extracts only the frequency components whose vibrations are to be suppressed using a Fourier transform. The control device then generates a compensation value for suppressing vibrations of the extracted frequency components through repetitive control using a repetitive compensator, and adds the compensation value to the motor command value. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Japanese Patent Application Laid-Open No. 2001-037287 Summary of the Invention [Problem to be solved by the invention]

[0004] In general, in various systems, there are cases where it is desired to extract only a frequency component of a specific frequency from an input signal containing multiple frequency components. As an example, in the motor system shown in Patent Document 1, only the frequency component of a frequency at which vibration is desired to be suppressed is extracted from a signal detected by an acceleration sensor. As a method for extracting only the frequency component of a specific frequency from an input signal, for example, a method using Fourier transform as shown in Patent Document 1 can be considered.

[0005] Specifically, a Fourier transformer extracts various frequency components from an input signal. Then, an inverse Fourier transformer restores a signal containing only frequency components of specific frequencies from the frequency components extracted by the Fourier transformer. However, such a method using a Fourier transform requires complex arithmetic processing including window functions and a large memory capacity for the calculations. As a result, the system may become complicated. On the other hand, a method using a bandpass filter may also be considered. However, a bandpass filter causes a phase delay, which complicates the design, especially when applied to a feedback system.

[0006] Therefore, a method using a tracking filter has been considered as a method for easily extracting only the frequency component of a specific frequency. However, since a tracking filter uses a low-pass filter to discriminate the specific frequency, it may extract not only the frequency component of the specific frequency but also frequency components nearby it. If nearby frequency components are extracted in this way, there is a risk of malfunction in a system equipped with the tracking filter, i.e., a system that performs predetermined processing using the output of the tracking filter.

[0007] The embodiments described below have been made in consideration of the above, and other problems and novel features will become apparent from the description of this specification and the accompanying drawings. [Means for solving the problem]

[0008] A semiconductor device according to one embodiment includes a tracking filter and a monitoring circuit that monitors the processing state of the tracking filter. The tracking filter extracts a frequency component of a tracking frequency from a tracking input signal containing multiple frequency components and generates a tracking output signal composed of the extracted frequency component of the tracking frequency. The tracking filter includes an extraction multiplier, a low-pass filter, and a restoration multiplier. The extraction multiplier multiplies the tracking input signal by a cosine wave signal or a sine wave signal of the tracking frequency. The low-pass filter removes frequency components higher than a cutoff frequency from the output signal of the extraction multiplier. The restoration multiplier generates the tracking output signal by multiplying the output signal of the low-pass filter by a cosine wave signal or a sine wave signal of the tracking frequency. Here, the monitoring circuit detects an AC component included in the output signal of the low-pass filter and determines whether or not there is an abnormality in the tracking filter based on the magnitude of the detected AC component. [Effects of the Invention]

[0009] According to the embodiment, it is possible to prevent malfunction of a system equipped with a tracking filter. [Brief explanation of the drawings]

[0010] [Figure 1] FIG. 1 is a block diagram showing an example of the configuration of a main part of a semiconductor device according to an embodiment. [Figure 2] FIG. 2 is a waveform diagram showing an ideal operation example of the tracking filter in FIG. [Figure 3] FIG. 3 is a schematic diagram illustrating an example of a problem that may occur in the tracking filter in FIG. [Figure 4A] FIG. 4A is a schematic diagram showing an example of an output signal when an ideal low-pass filter is used in FIG. [Figure 4B] FIG. 4B is a schematic diagram showing an example of an output signal when an actual low-pass filter is used in FIG. [Figure 5]FIG. 5 is a schematic diagram showing a configuration example of a motor system to which the semiconductor device according to one embodiment is applied. [Figure 6] FIG. 6 is a block diagram showing a detailed configuration example of the semiconductor device in FIG. [Figure 7A] FIG. 7A is a schematic diagram illustrating an example of torque vibration. [Figure 7B] FIG. 7B is a diagram showing an example of a frequency spectrum of the torque vibration shown in FIG. 7A. [Figure 8] FIG. 8 is a block diagram showing a detailed configuration example of the torque vibration compensator in FIG. [Figure 9] FIG. 9 is a block diagram showing a detailed configuration example of the vibration component extractor in FIG. [Figure 10] FIG. 10 is a schematic diagram showing an example of the general operation of the compensation value generating circuit in FIG. [Figure 11] FIG. 11 is a schematic diagram showing an example of the operation of the second determination circuit in FIG. 9 when the rotation speed of the motor is relatively high. [Figure 12] FIG. 12 is a schematic diagram showing an example of the operation of the second determination circuit in FIG. 9 when the rotation speed of the motor is relatively low. [Figure 13] FIG. 13 is a schematic diagram showing an example of the operation of the first determination circuit in FIG. [Figure 14] FIG. 14 is a schematic diagram showing an example of the configuration in which the torque vibration compensator shown in FIG. 8 is expanded. [Figure 15A] FIG. 15A is a flowchart showing an example of the processing contents of the main part of the torque vibration compensator shown in FIG. [Figure 15B] FIG. 15B is a flowchart showing an example of the processing content following FIG. 15A. [Figure 16] FIG. 16 is a schematic diagram illustrating an example of the processing content when determining whether or not the tracking output signal is in a steady state in FIG. 15A. [Figure 17] FIG. 17 is a schematic diagram showing an example of a configuration different from that shown in FIG. 14, which is an extension of the torque vibration compensator shown in FIG. [Figure 18A]FIG. 18A is a flowchart showing an example of the processing contents of the main part of the torque vibration compensator 108b shown in FIG. [Figure 18B] FIG. 18B is a flowchart showing an example of the processing content following FIG. 18A. DETAILED DESCRIPTION OF THE INVENTION

[0011] In the following embodiments, when necessary for convenience, the description will be divided into multiple sections or embodiments, but unless otherwise specified, they are not unrelated to each other, and one is a partial or complete modification, detail, supplementary explanation, etc. of the other. Furthermore, in the following embodiments, when the number of elements, etc. (including the number, numerical value, amount, range, etc.) is mentioned, it is not limited to that specific number, and may be more or less than the specific number, unless otherwise specified or when it is clearly limited in principle to a specific number.

[0012] Furthermore, in the following embodiments, it goes without saying that the components (including element steps, etc.) are not necessarily essential unless otherwise specified or considered to be clearly essential in principle. Similarly, in the following embodiments, when referring to the shape, positional relationship, etc. of components, etc., it is intended to include those that are substantially similar or similar to the shape, etc., unless otherwise specified or considered to be clearly not essential in principle. The same applies to the above numerical values ​​and ranges.

[0013] Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In all drawings for explaining the embodiments, the same components are generally designated by the same reference numerals, and repeated description thereof will be omitted.

[0014] <Outline of semiconductor device> Fig. 1 is a block diagram showing an example of the configuration of the main components of a semiconductor device according to one embodiment. The semiconductor device 10 shown in Fig. 1 is, for example, configured from a single semiconductor chip and provided as one component of a predetermined system. The semiconductor device 10 includes a tracking filter TF and a monitoring circuit MNI. The tracking filter TF and the monitoring circuit MNI are realized, for example, by dedicated hardware circuits or programmable logic devices such as FPGAs (Field Programmable Gate Arrays).

[0015] Alternatively, the tracking filter TF and the monitoring circuit MNI may be realized by software processing. In this case, the semiconductor device 10 includes a memory that stores a program and a processor that executes the program stored in the memory. The processor causes the semiconductor device 10 to function as the tracking filter TF and the monitoring circuit MNI based on the program.

[0016] <<Tracking filter details>> Fig. 2 is a waveform diagram showing an ideal operation example of the tracking filter TF in Fig. 1. Ideally, as shown in Fig. 2, the tracking filter TF selects a set tracking frequency ω from a tracking input signal Fin containing a plurality of frequency components. r Frequency component Fin(ω r ) is extracted. Then, the tracking filter TF extracts the extracted tracking frequency ω r Frequency component Fin(ω r ) and outputs a tracking output signal Fout.

[0017] 1, the tracking filter TF specifically includes extraction multipliers ML1x and ML1y, low-pass filters LPFx and LPFy, restoration multipliers ML2x and ML2y, and an adder ADD. The extraction multipliers ML1x and ML1y apply a tracking frequency ω r cosine wave signal (cos(ω rt)) or a sinusoidal signal (sin(ω r Specifically, the multiplier (first multiplier) ML1x multiplies a cosine wave signal (cos(ω r The multiplier (second multiplier) ML1y generates an output signal F1x by multiplying a sine wave signal (sin(ω r t)) to generate the output signal F1y.

[0018] The low-pass filters LPFx and LPFy may be first-order or second-order filters, or may be higher-order filters. When configured with digital circuits, the low-pass filters LPFx and LPFy may be realized by, for example, an IIR (Infinite Impulse Response) filter. The low-pass filters LPFx and LPFy pass frequency components lower than the cutoff frequency from the output signals of the extraction multipliers ML1x and ML1y, and remove frequency components higher than the cutoff frequency.

[0019] Specifically, a low-pass filter (first low-pass filter) LPFx receives the output signal F1x of the multiplier ML1x and generates a filtered output signal F2x, and a low-pass filter (second low-pass filter) LPFy receives the output signal F1y of the multiplier ML1y and generates a filtered output signal F2y.

[0020] The restoration multipliers ML2x and ML2y apply a tracking frequency ω r cosine wave signal (cos(ω r t)) or a sinusoidal signal (sin(ω r t)) to generate the tracking output signal Fout via the adder ADD. In detail, the multiplier ML2x multiplies the output signal F2x by a cosine wave signal (cos(ω r The multiplier ML2y multiplies the output signal F2y by a sine wave signal (sin(ω r Then, an adder ADD adds the output signals of the multipliers ML2x and ML2y to generate a tracking output signal Fout.

[0021] Here, the tracking filter TF will be described in further detail. First, the tracking input signal Fin is expressed by the formula (1) using a Fourier series. In the formula (1), "n" is an integer equal to or greater than 1. "ω n " represents a frequency that changes by an integer multiple, specifically an angular frequency. n " and "b n ” respectively at a certain frequency “ω n " are the Fourier cosine and sine coefficients in ". "A0" is the DC component.

number

[0022] The output signal F1x of the extraction multiplier ML1x is expressed by the formula (2). The low-pass filter LPFx is a low-frequency signal “ω n -ω r " frequency components pass through, while the high-frequency "ω n +ω r As a result, the output signal F2x of the low-pass filter LPFx is expressed by equation (3). In equation (3), ’ r ” is the tracking frequency ω r The frequency is in the vicinity of r ’ / 2” and “b r ’ / 2” are the pass frequencies “ω ’ r -ω r are the amplitudes of the cosine and sine components in

[0023]

number

number

[0024] Similarly, the output signal F1y of the extraction multiplier ML1y is expressed by the formula (4). The low-pass filter LPFy is a low-frequency signal "ω n -ω r " frequency components pass through, while the high-frequency "ω n +ω r As a result, the output signal F2y of the low-pass filter LPFy is expressed by equation (5), similarly to equation (3).

[0025]

number

number

[0026] Here, if the low-pass filters LPFx and LPFy are ideal, they pass only the DC component and remove the AC component. That is, the ideal low-pass filters LPFx and LPFy are expressed by adding "ω ’ r =ω r As a result, the output signals F2x and F2y of the ideal low-pass filters LPFx and LPFy are expressed by the DC components DCx and DCy shown in equation (6), respectively.

number

[0027] As a result, the output signals of the restoration multipliers ML2x and ML2y are expressed by equations (7) and (8), respectively. The tracking output signal Fout from the adder ADD is expressed by equation (9). That is, the tracking output signal Fout has an amplitude "a r / 2” with tracking frequency ω r and a cosine signal with amplitude “b r / 2” with tracking frequency ω r It is expressed as the sum of the sinusoidal signal.

[0028]

number

number

number

[0029] <<Problems with tracking filters>> 3 is a schematic diagram illustrating an example of a problem that may occur in the tracking filter TF in FIG. r In addition to the frequency components at the tracking frequency ω r It is assumed that the tracking frequency ω includes nearby frequency components. r As a nearby frequency component, a tracking frequency ω r harmonic components of the tracking frequency ω r Examples of distortion components that may be included in the cosine wave of

[0030] In the example shown in FIG. 3, the tracking frequency ω r Of the harmonic components of r ) are shown. In particular, the harmonic components of the tracking frequency ω r , that is, the lower the frequency to be extracted, the higher the tracking frequency ω r and the second frequency (2ω r ) are neighbors of each other.

[0031] Here, in an ideal low-pass filter LPF (Ideal), "ω ’ r =ω r In order to extract only the DC component obtained by substituting ", the cutoff frequency is set to approximately zero, and a steep filter characteristic can be formed. As a result, as shown in FIG. 3, the ideal low-pass filter LPF (Ideal) has a tracking frequency ω rOnly the frequency component of the second order (2ω r ) frequency components can be completely removed.

[0032] However, it is usually difficult to realize such a steep filter characteristic in an actual low-pass filter LPF(Actual). As a result, the actual low-pass filter LPF(Actual) has a gentle filter characteristic as shown in FIG. 3, and the tracking frequency ω r Not only the frequency component of r ) frequency components are also passed.

[0033] In this way, the second-order frequency (2ω r ) frequency component cannot be removed, the output signals F2x and F2y of the low-pass filters LPFx and LPFy are expressed by the formulas (10) and (11), respectively. In the formula (10), the output signal F2x includes the addition of "ω ’ r =ω r " is obtained by substituting the DC component DCx (=a r / 2), plus “ω ’ r =2ω r Similarly, with regard to equation (11), the output signal F2y includes an AC component ACy in addition to a DC component DCy.

[0034]

number

number

[0035] As a result, the output signals of the restoration multipliers ML2x and ML2y are expressed by equations (12) and (13), respectively. The tracking output signal Fout from the adder ADD is expressed by equation (14). As shown in equation (14), the second-order frequency (2ω rWhen the frequency component of the tracking output signal Fout cannot be removed, the tracking frequency ω r In addition to the frequency components of r ) frequency components. As a result, the tracking output signal Fout has a waveform that is similar to the ideal waveform shown in FIG.

[0036]

number

number

number

[0037] As described above, the tracking filter TF controls the tracking frequency ω r Not only the frequency component of , but also the nearby frequency components, for example, the second frequency (2ω r ) is extracted, a malfunction may occur in a system equipped with the tracking filter TF. For example, a system that performs a predetermined process using a tracking output signal Fout usually extracts a frequency component of the tracking frequency ω r Predetermined processing is performed on the assumption that only frequency components of the signal are included. If this assumption is not met, unintended processing results may occur. For this reason, the monitoring circuit MNI shown in Figure 1 is provided.

[0038] The tracking filter TF is not limited to the configuration example shown in Fig. 1 and can be modified as appropriate. For example, a configuration can be given in which extraction multipliers ML1x and ML1y are provided as part of a rotating coordinate converter that performs conversion to rotating coordinates, and restoration multipliers ML2x and ML2y are provided as part of a fixed coordinate converter that performs conversion to fixed coordinates. The tracking filter TF may include such a configuration as long as it has at least one of extraction multipliers ML1x and ML1y, one of low-pass filters LPFx and LPFy, and one of restoration multipliers ML2x and ML2y.

[0039] <<Monitoring circuit details>> 4A is a schematic diagram showing an example of an output signal F2x when an ideal low-pass filter LPFx is used in FIG. 1. The output signal F2x shown in FIG. 4A rises in a predetermined rise time based on the characteristics of the low-pass filter LPFx, and then becomes a DC component DCx (=a r Although not shown in the figure, the same is true for the output signal F2y.

[0040] 4B is a schematic diagram showing an example of an output signal F2x when an actual low-pass filter LPFx is used in FIG. 1. The output signal F2x shown in FIG. 4B rises in a predetermined rise time based on the characteristics of the low-pass filter LPFx, and then, as shown in equation (10), a DC component DCx (=a r / 2) and AC component ACx.

[0041] The magnitude (|ACx|) of the AC component ACx included in the output signal F2x, in other words, the amplitude, is expressed by equation (15) based on equation (10). Although not shown, the same applies to the output signal F2y. The magnitude (|ACy|) of the AC component ACy included in the output signal F2y, in other words, the amplitude, is also expressed by equation (15) based on equation (11).

number

[0042] Here, the monitoring circuit MNI shown in Fig. 1 monitors the processing state of the tracking filter TF. Schematically, the monitoring circuit MNI detects AC components ACx and ACy contained in the output signals F2x and F2y of the low-pass filters LPFx and LPFy. Then, the monitoring circuit MNI determines whether or not there is an abnormality in the tracking filter TF based on the magnitudes (|ACx|, |ACy|) of the detected AC components ACx and ACy. The monitoring circuit MNI notifies a predetermined system of the determination result of the presence or absence of an abnormality using an abnormality flag signal FLG.

[0043] In detail, the monitoring circuit MNI includes AC component detection circuits ADTx and ADTy, DC component detection circuits DDTx and DDTy, abnormality index calculation circuits IDXCx and IDXCy, and a judgment circuit JDG. The AC component detection circuit ADTx, DC component detection circuit DDTx, and abnormality index calculation circuit IDXCx process the output signal F2x of the low-pass filter LPFx. The AC component detection circuit ADTy, DC component detection circuit DDTy, and abnormality index calculation circuit IDXCy process the output signal F2y of the low-pass filter LPFy.

[0044] With respect to the output signal F2x, the AC component detection circuit ADTx and the DC component detection circuit DDTx respectively detect the AC component (first AC component) ACx and the DC component DCx contained in the output signal F2x. At this time, the DC component detection circuit DDTx detects the DC component DCx shown in FIG. 4B by, for example, processing the output signal F2x using a moving average filter or the like. The AC component detection circuit ADTx detects the AC component ACx shown in FIG. 4B, specifically its magnitude (|ACx|), by, for example, monitoring the peak value of the output signal F2x. Similarly, with respect to the output signal F2y, the AC component detection circuit ADTy and the DC component detection circuit DDTy respectively detect the AC component (second AC component) ACy and the DC component DCy contained in the output signal F2y.

[0045] The abnormality index calculation circuit IDXCx calculates the ratio of the AC component ACx to the DC component DCx, specifically the magnitude (|ACx|), as the abnormality index IDXx, as shown in equation (16). Similarly, the abnormality index calculation circuit IDXCy calculates the ratio of the AC component ACy to the DC component DCy, specifically the magnitude (|ACy|), as the abnormality index IDXy, as shown in equation (16). Note that the DC components DCx and DCy may be negative values. For this reason, the abnormality indexes IDXx and IDXy may be absolute values ​​of those calculated by equation (16).

number

[0046] The determination circuit JDG generally determines whether or not there is an abnormality in the tracking filter TF based on the magnitude of at least one of the AC component ACx and the AC component ACy. Specifically, the determination circuit JDG determines whether or not there is an abnormality when the abnormality indexes IDXx and IDXy are greater than predetermined thresholds (first thresholds) ITHx and IThy, respectively, and determines whether or not there is an abnormality when the abnormality indexes IDXx and IDXy are smaller than the thresholds ITHx and IThy. In this case, the determination circuit JDG may perform the determination using an AND condition or an OR condition for the two abnormality indexes IDXx and IDXy. The determination circuit JDG then notifies the system of the determination result of whether or not there is an abnormality using an abnormality flag signal FLG.

[0047] The monitoring circuit MNI may monitor only one of the output signals F2x and F2y. That is, the monitoring circuit MNI may be configured without the AC component detection circuit ADTy, the DC component detection circuit DDTy, and the abnormality index calculation circuit IDXCy. However, in order to perform monitoring with higher reliability, it is preferable that the monitoring circuit MNI monitors both the output signals F2x and F2y.

[0048] By providing the monitoring circuit MNI as described above, it is possible to prevent malfunction of a system equipped with the tracking filter TF. That is, the system, for example, r If a predetermined process is performed based on the tracking output signal Fout that contains a predetermined amount or more of frequency components of the above frequency band, an unintended processing result may be generated. Therefore, if the tracking output signal Fout is deemed to be abnormal, an abnormality flag signal FLG is used to notify the system. This makes it possible to prevent a situation in which an unintended processing result is generated.

[0049] The monitoring circuit MNI may determine the presence or absence of an abnormality simply based on the magnitude (|ACx|) of the AC component ACx, instead of the abnormality index IDXx representing the ratio (|ACx| / DCx) as shown in FIG. 1. However, the magnitude (|ACx|) of the AC component ACx may change as appropriate depending on, for example, the system specifications, environment, etc. For example, in a motor system, the average magnitude (|ACx|) of the AC component, i.e., the range, may change depending on the motor capacity, the specifications of the load driven by the motor, etc. As a result, it becomes necessary to set multiple threshold values ​​ITHx depending on the system specifications, environment, etc., which may result in complex settings and reduced versatility.

[0050] From this perspective, it is more beneficial to use the abnormality index IDXx instead of the magnitude (|ACx|) of the AC component ACx. That is, in this case, the monitoring circuit MNI can determine whether or not there is an abnormality based on the magnitude (|ACx|) of the AC component relative to the DC component DCx. This allows a single threshold value ITHx to be used for general purposes. For example, in a motor system, the monitoring circuit MNI can discriminate the motor capacity using the DC component DCx, and can determine whether or not there is an abnormality based on the magnitude (|ACx|) of the AC component normalized by the discriminated motor capacity. As a result, the same threshold value ITHx can be applied even if the motor capacity changes, for example.

[0051] <Application example to motor systems> FIG. 5 is a schematic diagram showing an example of the configuration of a motor system to which a semiconductor device according to one embodiment is applied. The motor system shown in FIG. 5 includes a semiconductor device 10a, an inverter 20, and a motor MT. The motor MT is, for example, a three-phase motor consisting of a u-phase, a v-phase, and a w-phase. The semiconductor device 10a generates and outputs a motor control signal for controlling the motor MT, for example, a PWM (Pulse Width Modulation) signal Gpwm. The inverter 20 supplies AC power to each phase of the motor MT based on the motor control signal. As will be described in detail later, the example configuration shown in FIG. 1 is applied to a part of the semiconductor device 10a.

[0052] In this example, the semiconductor device 10a is a microcontroller or SoC (System on Chip) configured on a single semiconductor chip. The semiconductor device 10a mainly includes a processor PRC, a memory MEM, a PWM signal generator PWMG, and an analog-to-digital converter ADC. The processor PRC is, for example, a CPU (Central Processing Unit) or a DSP (Digital Signal Processor). The memory MEM includes a ROM (Read Only Memory) and a RAM (Random Access Memory). The ROM is, for example, a flash memory. The RAM is, for example, an SRAM or a DRAM.

[0053] The ROM stores a motor control program. The motor control program is copied to the RAM. The processor PRC controls the motor MT by executing the motor control program copied to the RAM. The PWM signal generator PWMG generates a PWM signal Gpwm(u,v,w) for each phase based on, for example, a duty ratio command value for each phase from the processor PRC. The analog-to-digital converter ADC receives a detection signal representing the state of the motor MT, in this example, a detection signal representing the phase current of the motor MT, via the inverter 20. The analog-to-digital converter ADC converts the detection signal into a digital value.

[0054] The inverter 20 includes a gate driver GD, a switching circuit SWC, and a current detector IDET. The switching circuit SWC is configured, for example, as a three-phase bridge circuit consisting of six switching elements. The gate driver GD receives PWM signals Gpwm(u,v,w) for each phase from the semiconductor device 10 and controls the on / off of the six switching elements in the switching circuit SWC based on the PWM signals Gpwm(u,v,w). As a result, the switching circuit SWC supplies three-phase voltages Vu, Vv, and Vw to the motor MT according to the duty ratios of the PWM signals.

[0055] The current detector IDET detects phase currents Iu and Iw flowing through at least two phases of the motor MT (in this example, the u-phase and w-phase). An analog-to-digital converter ADC in the semiconductor device 10 converts the phase currents Iu and Iw detected by the current detector IDET into digital values. The phase current Iv flowing through the remaining phase can be calculated using the relationship "Iu + Iv + Iw = 0." Alternatively, a method may be used in which the three phase currents Iu, Iv, and Iw are combined and detected by a single current detector. In this case, the current detector recognizes the on / off states of six switching elements based on, for example, the PWM signal Gpwm(u, v, w) and detects the corresponding phase current based on the combination of the on / off states.

[0056] Fig. 6 is a block diagram showing a detailed example configuration of semiconductor device 10a in Fig. 5. Semiconductor device 10a shown in Fig. 6 includes a motor controller 100 in addition to the PWM signal generator PWMG, analog-to-digital converter ADC, RAM, and ROM shown in Fig. 5. Motor controller 100 is realized by processor PRC shown in Fig. 5 executing a motor control program stored in RAM. In other words, the motor control program causes processor PRC to function as each component within motor controller 100 shown in Fig. 6.

[0057] The motor controller 100 includes a speed commander 101, a speed controller 102, a current controller 103, a two-axis to three-axis converter 104, a PWM signal modulator 105, a three-axis to two-axis converter 106, a rotation angle / speed estimator 107, a torque vibration compensator 108, and an adder 109.

[0058] The three-axis / two-axis converter 106 receives two-phase currents Iu and Iw from the analog-digital converter ADC and the rotation angle θ from the rotation angle / speed estimator 107. Then, the three-axis / two-axis converter 106 converts the three-phase currents Iu, Iv, and Iw in the UVW coordinate system obtained from the two-phase currents Iu and Iw into a d-axis current Id and a q-axis current Iq in the dq coordinate system by Clarke transformation and Park transformation using the rotation angle θ. The UVW coordinate system is a rotating coordinate system. On the other hand, the dq coordinate system is a fixed coordinate system.

[0059] The speed commander 101 generates a speed command value ω based on a predetermined speed profile, for example. * The speed controller 102 generates the speed command value ω * and the value of the rotational speed ω from the rotation angle / speed estimator 107, the speed controller 102 performs, for example, PI (proportional-integral) control. * , in other words, generates a torque command value.

[0060] The adder 109 calculates the q-axis current command value Iq from the speed controller 102. * and the q-axis current command value Iq from the torque vibration compensator 108. ** The sum is used as the q-axis current command value Iq *** The torque vibration compensator 108 outputs, for example, a speed command value ω * and the value of the rotational speed ω from the rotation angle / speed estimator 107, a torque compensation value for suppressing torque vibration is calculated as a q-axis current command value Iq ** Accordingly, an adder 109 adds a torque compensation value to the torque command value.

[0061] The current controller 103 receives the q-axis current command value Iq from the adder 109. *** and the d-axis current command value Id * Input the d-axis current command value Id * is fixed to zero, for example. However, the d-axis current command value Id * may be a non-zero value generated based on the flux-weakening control or the "Maximum Torque Per Ampere" control. * and the q-axis current command value Iq *** and the d-axis current Id and the q-axis current Iq from the 3-axis / 2-axis converter 106. Based on this, the current controller 103 performs PI control or the like. * and the q-axis voltage command value Vq* Generate.

[0062] The 2-axis / 3-axis converter 104 converts the d-axis voltage command value Vd * and the q-axis voltage command value Vq * and the rotation angle θ from the rotation angle / speed estimator 107. At this time, the rotation angle θ may be corrected in consideration of the rotation of the motor MT. The 2-axis / 3-axis converter 104 performs an inverse Park transformation and an inverse Clarke transformation using the rotation angle θ to derive a d-axis voltage command value Vd * and the q-axis voltage command value Vq * is the three-phase voltage command value Vu * ,Vv * ,Vw * The PWM signal modulator 105 converts the three-phase voltage command value Vu * ,Vv * ,Vw * are converted into three-phase duty ratio command values ​​Du, Dv, Dw and output to the PWM signal generator PWMG.

[0063] The rotation angle / speed estimator 107 calculates the d-axis current Id and the q-axis current Iq from the 3-axis / 2-axis converter 106 and the d-axis voltage command value Vd from the current controller 103. * and the q-axis voltage command value Vq * and are input. The rotation angle / speed estimator 107 calculates the d-axis and q-axis induced voltages based on these input values ​​and a predetermined state equation of the motor. The rotation angle / speed estimator 107 then calculates, in other words, estimates or detects, the rotation angle θ based on the calculated d-axis and q-axis induced voltages. Furthermore, the rotation angle / speed estimator 107 calculates the rotation speed ω by performing a differential operation on the rotation angle θ. Note that the rotation angle / speed estimator 107 may calculate the magnetic flux instead of the induced voltage to calculate the rotation angle θ.

[0064] FIG. 6 illustrates a motor controller 100 that performs sensorless vector control. However, the motor controller 100 may also perform sensor-equipped vector control. In this case, a position / speed sensor is installed in the motor MT instead of the rotation angle / speed estimator 107. The position / speed sensor is, for example, a rotary encoder that detects the rotation angle θ and the rotation speed ω. Here, the motor controller 100 is implemented by program processing using a processor PRC. However, the motor controller 100 may also be implemented using, for example, an FPGA (Field Programmable Gate Array) or an ASIC. That is, the semiconductor device 10a shown in FIG. 6 may be an FPGA, an ASIC, or the like.

[0065] <About torque vibration> FIG. 7A is a schematic diagram illustrating an example of torque vibration. For example, in FIGS. 5 and 6, when torque is applied to the motor MT as a disturbance, the rotation speed ω of the motor MT vibrates. This type of vibration is called torque vibration. As a specific example, as shown in FIG. 7A, a compressor motor installed in an air conditioner sequentially expands, draws in, compresses, and exhausts the refrigerant within a period in which the motor MT, or more specifically, the rotor, rotates through a mechanical angle of 360 degrees. This causes the load torque to fluctuate. Torque vibration can then occur due to the difference between the load torque and the output torque.

[0066] This torque vibration occurs periodically depending on the rotational angle (mechanical angle) of the motor MT, in other words, the magnetic pole position, and changes the rotational speed ω of the motor MT. Furthermore, the frequency of this torque vibration is synchronized with the rotational speed ω of the motor MT. Therefore, low-frequency torque vibration occurs at low speeds. In particular, low-frequency torque vibration not only reduces controllability, but also generates noise and shortens the lifespan of the motor system. For this reason, it is desirable to suppress torque vibration.

[0067] Fig. 7B is a diagram showing an example of the frequency spectrum of the torque vibration shown in Fig. 7A. As shown in Fig. 7B, the torque vibration waveform shown in Fig. 7A is decomposed into first-, second-, third-, and other frequency components. As shown in Fig. 7A, the first-order frequency component is a vibration component at a fundamental frequency whose cycle is a time period corresponding to a mechanical angle of 360 degrees. The second-, third-, and other frequency components are vibration components at frequencies twice, three times, and so on, the fundamental frequency, and represent distortion components superimposed on the fundamental frequency waveform in the torque vibration waveform shown in Fig. 7A.

[0068] <Details of the torque vibration compensator> Fig. 8 is a block diagram showing a detailed example configuration of the torque vibration compensator 108 in Fig. 6. The torque vibration compensator 108 shown in Fig. 8 includes a rotation angle converter 120, a vibration component extractor 121, and a compensation value generation circuit 125. The rotation angle converter 120 converts the rotation angle θ as an electrical angle from the rotation angle / speed estimator 107 shown in Fig. 6 into a mechanical angle θm based on the number of poles of the motor MT.

[0069] The vibration component extractor 121 includes a tracking filter TFa and a monitoring circuit MNIa as shown in FIG. 1, and a multiplier 124 that multiplies by a compensation value gain k1. As will be described in detail later, the vibration component extractor 121 uses the tracking filter TFa to extract the torque vibration component as described in FIG. 7A and FIG. 7B, more specifically, a cancellation component for canceling the torque vibration component, as a tracking output signal Fout. The vibration component extractor 121 then multiplies the tracking output signal Fout by the compensation value gain k1 to generate an update amount UA of the torque compensation value TCV for canceling the torque vibration component, and outputs the generated update amount UA to the compensation value generation circuit 125.

[0070] In this example, the compensation value generation circuit 125 includes a compensation value table update circuit 122 and a compensation value interpolation circuit 123. The compensation value generation circuit 125 also stores a compensation value table CTBL in a memory MEM, specifically a RAM. As will be described in detail later, the compensation value generation circuit 125 generates a torque compensation value TCV for suppressing torque vibration while successively updating it through a learning operation using an update amount UA based on the tracking output signal Fout as an input. The compensation value generation circuit 125 then converts the generated torque compensation value TCV into a q-axis current command value Iq for vibration suppression shown in FIG. 6. ** By outputting it as such, it is reflected in the PWM signal Gpwm, which is the motor control signal.

[0071] <<More details about the vibration component extractor>> Fig. 9 is a block diagram showing a detailed configuration example of the vibration component extractor 121 in Fig. 8. In Fig. 9, the tracking filter TFa has the same configuration as in Fig. 1. However, in this example, the tracking filter TFa converts the speed command value ω * and the value of the rotational speed ω from the rotation angle / speed estimator 107 is input as a tracking input signal Fin. As a result, the tracking input signal Fin is calculated based on the speed command value ω as shown in FIG. * It represents the torque vibration, which is also the change in the rotation speed ω based on the reference.

[0072] In detail, the tracking filter TFa is * 7A, the tracking input signal Fin is obtained by inputting the values ​​of the speed command value ω and the rotation speed ω as positive and negative inputs, respectively. As a result, the tracking input signal Fin becomes a signal that cancels out the speed deviation, and for example, in FIG. * The signal is linearly symmetric with the signal of the rotation speed ω, with the center line at . By using such a speed deviation as the tracking input signal Fin, for example, an acceleration sensor as shown in Patent Document 1 becomes unnecessary, thereby reducing costs.

[0073] The tracking filter TFa also receives an input of the mechanical angle θm from the rotation angle converter 120. The tracking filter TFa multiplies the input mechanical angle θm by the "ω" of each of the multipliers ML1x, ML1y, ML2x, and ML2y shown in FIG. r t”, i.e., the tracking frequency ω r As a result, the tracking filter TFa applies the fundamental frequency (first order) corresponding to one period of the mechanical angle θm described in FIGS. 7A and 7B to the tracking frequency ω r and the frequency components can be extracted.

[0074] Then, tracking filter TFa ideally generates a tracking output signal Fout composed of the extracted frequency components, here fundamental frequency (first-order) frequency components. This tracking output signal Fout is a signal that cancels out the fundamental frequency (first-order) frequency component contained in the detected torque vibration. Note that tracking filter TFa can also extract second-order, third-order, etc. frequency components by doubling, tripling, etc. the input mechanical angle θm, and can also generate corresponding tracking output signals Fout for each.

[0075] The monitoring circuit MNIa includes a second determination circuit JDG2 and an OR operation circuit OR in addition to the configuration shown in Fig. 1. The second determination circuit JDG2 detects the amplitude of the tracking output signal Fout by envelope detection or the like, and determines whether suppression of torque vibration has been completed or not based on the magnitude of the detected output amplitude. The second determination circuit JDG2 then outputs a suppression completion signal SC that indicates the determination result.

[0076] In detail, the second decision circuit JDG2 first determines the magnitude of the output amplitude of the tracking output signal Fout detected at the start of the learning operation performed by the compensation value generation circuit 125 as the pre-suppression amplitude. Then, while sequentially detecting the output amplitude during the learning operation, the second decision circuit JDG2 sequentially calculates the ratio of the detected output amplitude to the pre-suppression amplitude as the suppression rate. If the calculated suppression rate is lower than a predetermined threshold value (second threshold value) FTH, the second decision circuit JDG2 determines that suppression is complete, and if it is higher than the threshold value FTH, it determines that suppression is incomplete.

[0077] The OR circuit OR outputs, or asserts, the learning completion signal LCP when at least one of the suppression completion signal SC and the abnormality flag signal FLG is output, or in other words, asserted. That is, the monitoring circuit MNIa outputs the learning completion signal LCP via the OR circuit OR in a first case or a second case. The first case is when it is determined that suppression of torque vibration is completed based on the suppression completion signal SC. The second case is when it is determined that there is an abnormality in the tracking filter TFa based on the abnormality flag signal FLG.

[0078] <<Details of the compensation value generation circuit>> FIG. 10 is a schematic diagram showing an example of the general operation of the compensation value generation circuit 125 in FIG. 8. In FIG. 10, a compensation value table CTBL registers rotation angles, which are discrete values, more specifically, torque compensation values ​​TCV for each mechanical angle θm. In the specification, the discretized rotation angle of the motor MT registered in the compensation value table CTBL is also referred to as the discretized rotation angle θm. In this example, the discretized rotation angle θm is represented by 64 index numbers NUM[0], [1], ...,

[63] , which correspond to one rotation. In this case, the mechanical angle θm between adjacent index numbers NUM, i.e., the resolution, is 5.625 (=360 / 64) [deg].

[0079] As a result, for example, the discretized rotation angles θm corresponding to the first index number NUM[0] and the last index number NUM

[63] are 0 [deg] and 354.375 [deg], respectively. Here, because repeated control is performed at a period of mechanical angle θm of 360 [deg], the index number NUM next to the last index number NUM

[63] is the first index number NUM[0]. The compensation value table update circuit 122 sequentially updates this compensation value table CTBL based on the update amount UA from the vibration component extractor 121. That is, the compensation value table CTBL is sequentially updated by a learning operation.

[0080] In detail, for example, when the mechanical angle θm of the motor MT reaches a certain discretized rotation angle θm, the vibration component extractor 121 performs arithmetic processing using the discretized rotation angle θm, and outputs the update amount UA corresponding to the discretized rotation angle θm. As a specific example, assume that, at the start of learning shown in Fig. 10, the mechanical angle θm reaches the discretized rotation angle θm corresponding to index number NUM[2], i.e., 11.25 [deg].

[0081] In this case, the tracking filter TFa outputs an offset value Fout[2] at 11.25[deg] in the tracking output signal Fout for offsetting the fundamental frequency (first-order) frequency component included in the torque vibration. The vibration component extractor 121 multiplies the offset value Fout[2] by the compensation value gain k1 to output the update amount UA[2] of "0.01".

[0082] In response to this, the compensation value table update circuit 122 adds the update amount UA[2] of "0.01" to the current torque compensation value TCV of "0.0" at index number NUM[2] in the compensation value table CTBL. The same processing as for index number NUM[2] is performed for subsequent index numbers NUM[3], [4], .... In this way, the compensation value table update circuit 122 updates the compensation value table CTBL by integrating the update amount UA for each discretized rotation angle θm for each discretized rotation angle θm, i.e., for each index number NUM.

[0083] On the other hand, the compensation value interpolation circuit 123 sequentially calculates an interpolation function CF that represents the relationship between the mechanical angle θm and the torque compensation value TCV based on the compensation value table CTBL that is sequentially updated. The compensation value interpolation circuit 123 then calculates the torque compensation value TCV by substituting the input mechanical angle θm into the interpolation function CF, and calculates the q-axis current command value Iq for vibration suppression shown in FIG. ** As a result, the torque compensation value TCV is reflected in the PWM signal Gpwm, which is the motor control signal.

[0084] Here, the tracking output signal Fout essentially represents a cancellation component corresponding to the torque vibration component that still remains after the current torque compensation value TCV is reflected in the control of the motor MT. Therefore, as shown in Fig. 10, the amplitude of the tracking output signal Fout decreases as the learning operation of the compensation value table CTBL progresses and the suppression of the torque vibration component progresses accordingly. Accordingly, the torque compensation value TCV registered in the compensation value table CTBL converges to a predetermined value.

[0085] Furthermore, in order to reduce the required memory capacity, the compensation value table CTBL registers the torque compensation value TCV for each discretized rotation angle θm. However, the compensation value interpolation circuit 123 needs to input the mechanical angle θm, which changes almost continuously, and output the torque compensation value TCV. Therefore, as shown in FIG. 10, the compensation value interpolation circuit 123 calculates an interpolation function CF for converting the discrete value into a continuous value. The interpolation function CF may be, for example, a function obtained by performing polynomial approximation on the torque compensation values ​​TCV for all discretized rotation angles θm. Alternatively, the interpolation function CF may be, for example, a linear or quadratic interpolation function that interpolates between two adjacent torque compensation values ​​TCV.

[0086] The compensation value generation circuit 125 may be any circuit that repeatedly generates the torque compensation value TCV by performing learning operations such as integral compensation and proportional integral compensation using the tracking output signal Fout as an input, and is not particularly limited to the configuration shown in Fig. 8. For example, the compensation value generation circuit 125 may be a repetitive compensator such as that shown in Patent Document 1.

[0087] <<About the learning completion signal>> Fig. 11 is a schematic diagram showing an example of the operation of the second judgment circuit JDG2 in Fig. 9 when the motor rotation speed is relatively high. In Fig. 11, for example, the output signal F2x of the low-pass filter LPFx rises after a predetermined control delay, and then gradually decreases as the learning operation in the compensation value generation circuit 125 and, ultimately, the suppression of torque vibration progresses. The same is true for the output signal F2y of the low-pass filter LPFy.

[0088] Here, when the rotation speed ω of the motor MT is medium or high, for example, in FIG. r " and "2ω r 11, the output signal F2x does not include the AC component ACx as shown in FIG. 4B, but only the DC component DCx. Based on this output signal F2x, the tracking output signal Fout becomes a signal that gradually attenuates as the learning operation progresses.

[0089] 11, the second decision circuit JDG2 determines the magnitude of the output amplitude (|Fout|) of the tracking output signal Fout detected at the start of the learning operation as the pre-suppression amplitude A[0]. Then, while sequentially detecting the output amplitude A[n] during the learning operation, the second decision circuit JDG2 sequentially calculates the ratio of the detected output amplitude A[n] to the pre-suppression amplitude A[0] as the suppression rate SR (=A[n] / A[0]). If the suppression rate SR is lower than a threshold value FTH, the second decision circuit JDG2 determines that suppression is complete, and if it is higher than the threshold value FTH, it determines that suppression is incomplete.

[0090] 11, the suppression rate SR is lower than the threshold value FTH at a certain point in time. Therefore, the second judgment circuit JDG2 determines that suppression is complete and outputs a suppression completion signal SC. As a result, the monitoring circuit MNIa outputs a learning completion signal LCP via the OR operation circuit OR.

[0091] Meanwhile, the compensation value generation circuit 125 shown in Fig. 8 completes the learning operation as described in Fig. 10 in response to the learning completion signal LCP. In the example shown in Fig. 10, the compensation value generation circuit 125 completes the learning operation by stopping the update operation of the compensation value table CTBL. Then, the compensation value generation circuit 125 continuously reflects the torque compensation value TCV at the time of completing the learning operation, i.e., the torque compensation value TCV based on the compensation value table CTBL at the time of completing the learning operation, in the motor control signal.

[0092] In a motor system, regular torque vibrations may occur steadily. Therefore, the torque vibration compensator 108 suppresses the torque vibrations by generating a torque compensation value TCV through a learning operation to offset the torque vibrations, thereby establishing a stable state in which the torque vibrations are sufficiently suppressed. After that, the torque vibration compensator 108 completes the learning operation and maintains the stable state by continuously reflecting the torque compensation value TCV at that time in the motor control signal. In the stable state, the update amount UA of the torque compensation value TCV may become zero, so there is no need to perform unnecessary learning operations.

[0093] 12 is a schematic diagram showing an example of the operation of the second determination circuit JDG2 when the rotation speed of the motor in FIG. 9 is relatively low. When the rotation speed ω of the motor MT is low, for example, in FIG. r " and "2ω r 11, the output signal F2x of the low-pass filter LPFx becomes a signal in which an AC component ACx as shown in FIG. 4B is superimposed on a DC component DCx that gradually decreases, as shown in FIG. 12.

[0094] Here, the motor system is driven via the torque vibration compensator 108 at a tracking frequency ω r A negative feedback system is formed to suppress the torque vibration of the tracking frequency ω r The frequency component of the first order (ω r ) frequency components are suppressed as the learning process progresses.

[0095] On the other hand, for example, the second order (2ω r ) frequency components, the first order (ω r ) frequency component, the phase characteristics of the feedback system are different from those of the feedback system. Therefore, the motor system does not necessarily form a negative feedback system, and in some cases may form a positive feedback system. As a result, the second-order (2ω r ) frequency components are not necessarily suppressed as the learning operation progresses, and may be amplified in some cases.

[0096] For this reason, the tracking output signal Fout becomes quadratic (2ω r ), that is, the influence of the AC component ACx, becomes relatively large. Accordingly, even if the learning operation progresses, the magnitude of the output amplitude of the tracking output signal Fout (|Fout|) decreases only to a certain extent due to the influence of the AC component ACx, and in some cases it may even increase.

[0097] As a result, the following problems may occur. First, as shown in FIG. 12, the suppression rate SR does not become lower than the threshold value FTH, so the suppression completion signal SC is not output and the learning operation is not completed. If the threshold value FTH is set high as a countermeasure, the suppression of torque vibrations becomes insufficient. Second, erroneous learning may occur. For example, the second order (2ω r ) as a new torque vibration source, a learning operation may be performed to maintain or amplify the new torque vibration source. If such a learning operation continues, it may cause an unexpected abnormality such as an overcurrent.

[0098] Therefore, an abnormality flag signal FLG from the judgment circuit (first judgment circuit) JDG shown in Fig. 9 is used in combination. Fig. 13 is a schematic diagram showing an example of the operation of the first judgment circuit JDG in Fig. 9. In Fig. 13, the output signal F2x of the low-pass filter LPFx is a signal in which an AC component ACx is superimposed on a DC component DCx that gradually decreases, as in the case of Fig. 12. The abnormality index calculation circuit IDXCx calculates the ratio of the AC component ACx to the DC component DCx, specifically, its magnitude (|ACx|), as the abnormality index IDXx.

[0099] As a result, the abnormality index IDXx increases as the learning operation progresses, as shown in FIG. 13. The first determination circuit JDG determines that an abnormality exists when the abnormality index IDXx is higher than a predetermined threshold (first threshold) ITHx, and determines that an abnormality does not exist when the abnormality index IDXx is lower than the threshold ITHx. Therefore, as shown in FIG. 13, the first determination circuit JDG determines that an abnormality exists at a predetermined point in time after the learning operation has progressed, and outputs an abnormality flag signal FLG. The monitoring circuit MNIa outputs a learning completion signal LCP in response to the abnormality flag signal FLG. The compensation value generation circuit 125 completes the learning operation in response to the learning completion signal LCP, and continuously reflects the torque compensation value TCV at the time of completion of the learning operation in the motor control signal.

[0100] In this way, by using the suppression completion signal SC and the abnormality flag signal FLG together, even if the suppression completion signal SC is not output, the primary (ω r The learning operation is completed when the suppression of the frequency component of (ω) has progressed to a certain extent. This prevents the progression of erroneous learning as described in FIG. 12 and the occurrence of unexpected abnormalities such as overcurrent. Furthermore, the learning operation can be effectively utilized in various rotation speed ω ranges. That is, the completion of the learning operation can be determined by the suppression completion signal SC in the medium or high speed range, and by the abnormality flag signal FLG in the low speed range.

[0101] <About the torque vibration compensator (application example [1])> Fig. 14 is a schematic diagram showing an example of the configuration obtained by expanding the torque vibration compensator shown in Fig. 8. In the torque vibration compensator 108a shown in Fig. 14, the example of the configuration in Fig. 8 is expanded to provide a plurality of tracking frequencies. In this example, a first-order tracking frequency ω r In addition to the second-order tracking frequency (2ω r Accordingly, the torque vibration compensator 108a includes two changeover switches SW1 and SW2 and two compensation value generation circuits 125-1 and 125-2.

[0102] The selector switch SW1 selects either the mechanical angle θm or the doubled mechanical angle (2θm) based on a first learning completion signal LCP1 from the monitoring circuit MNIa and outputs the selected angle to the tracking filter TFa. The doubled mechanical angle (2θm) is generated, for example, by the rotation angle converter 120 shown in FIG. 8. The selector switch SW2 outputs the update amount UA from the multiplier 124, here the first update amount UA1 or the second update amount UA2, to one of the two compensation value generation circuits 125-1, 125-2 based on the first learning completion signal LCP1.

[0103] In summary, the tracking filter TFa generates two tracking output signals Fout1 and Fout2 in a time-division manner by switching between the mechanical angle θm and the double mechanical angle (2θm) via the changeover switch SW1. In other words, the tracking filter TFa generates a first-order tracking frequency ω r and the second-order tracking frequency (2ω r ) to generate two tracking output signals Fout1 and Fout2 in a time-division manner.

[0104] In detail, the tracking filter TFa first inputs the mechanical angle θm to generate the first-order tracking frequency ω r Then, in response to the learning completion signal LCP1 from the monitoring circuit MNIa, the tracking filter TFa inputs a doubled mechanical angle (2θm) instead of the mechanical angle θm. This causes the tracking filter TFa to output a second-order tracking frequency (2ω r ) to generate a second tracking output signal Fout2. The multiplier 124 receives the first tracking output signal Fout1 and the second tracking output signal Fout2 and generates a first update amount UA1 and a second update amount UA2, respectively.

[0105] The compensation value generation circuit 125-1 receives the first update amount UA1 from the selector switch SW2 and, consequently, the first tracking output signal Fout1 as inputs, and starts a first learning operation using the first compensation value table CTBL1. More specifically, the compensation value generation circuit 125-1 starts the first learning operation in response to a first learning start signal LST1 from the monitoring circuit MNIa. Thereafter, the compensation value generation circuit 125-1 completes the first learning operation in response to a first learning completion signal LCP1, and continuously reflects the first torque compensation value TCV1 at the time of completion of the learning operation in the motor control signal.

[0106] Meanwhile, in response to the first learning completion signal LCP1, the selector switch SW2 switches the output destination from the compensation value generation circuit 125-1 to the compensation value generation circuit 125-2. The compensation value generation circuit 125-2 receives the second update amount UA2 from the selector switch SW2 and, in turn, the second tracking output signal Fout2, and starts a second learning operation using the second compensation value table CTBL2. More specifically, the compensation value generation circuit 125-2 starts the second learning operation in response to the second learning start signal LST2 from the monitoring circuit MNIa.

[0107] Thereafter, the compensation value generation circuit 125-2 completes the second learning operation in response to the second learning completion signal LCP2 from the monitoring circuit MNIa, and continuously reflects the second torque compensation value TCV2 at the time of completion of the learning operation in the motor control signal. The second torque compensation value TCV2 is a value that changes at twice the frequency of the first torque compensation value TCV1. The first learning start signal LST1 and the second learning start signal LST2 will be described in detail later.

[0108] 8, for example, the compensation value table update circuit 122 may be shared by the two compensation value generation circuits 125-1 and 125-2, but the compensation value table CTBL and the compensation value interpolation circuit 123 are provided separately for the two compensation value generation circuits 125-1 and 125-2.

[0109] In the example shown in FIG. 14, the two torque compensation values ​​TCV1 and TCV2 are added together after passing through the lead compensators 130-1 and 130-2, respectively, and output as the torque compensation value TCV. For example, when the compensation value generation circuit 125-1 outputs a first torque compensation value TCV1 corresponding to a certain mechanical angle θm, the first torque compensation value TCV1 is reflected in the motor MT after a predetermined control delay. However, during this control delay, the actual mechanical angle θm of the motor MT may advance, causing a deviation in the mechanical angle θm. The lead compensators 130-1 and 130-2 compensate for this deviation. Note that the lead compensators may also be applied to the configuration example shown in FIG. 8.

[0110] By using the above-described configuration example, it is possible to suppress the first and second frequency components contained in the torque vibration as shown in FIG. 7B. In this case, if the learning completion signal LCP is generated based only on the suppression rate SR as shown in FIG. 12, the first learning completion signal LCP1 in FIG. 14 is not output, and therefore the second tracking frequency (2ω r ) is not switched to. Therefore, if the abnormality index IDXx shown in Figure 13 is used in combination, the second-order tracking frequency (2ω r ), which results in the suppression of the second-order frequency components contained in the torque vibration.

[0111] Fig. 15A is a flow chart showing an example of the processing contents of the main parts of the torque vibration compensator 108a shown in Fig. 14. Fig. 15B is a flow chart showing an example of the processing contents subsequent to Fig. 15A. For example, the processor PRC in Fig. 5 executes these processes based on a program in the memory MEM. Fig. 16 is a schematic diagram for explaining an example of the processing contents when determining whether the tracking output signal is in a steady state in Fig. 15A.

[0112] 15A, steps S101-S104 are processes for preparing for the first learning operation. Steps S105-S107 are processes associated with the execution of the first learning operation. In step S101, the torque vibration compensator 108a uses the changeover switches SW1 and SW2 to change the mechanical angle θm, that is, the first-order tracking frequency ω r Accordingly, the tracking filter TFa selects the first-order tracking frequency ω r By extracting the frequency component of the signal, a first tracking output signal Fout1 is generated (step S102).

[0113] Next, the torque vibration compensator 108a waits for the first tracking output signal Fout1 to reach a steady state (step S103). That is, the first tracking output signal Fout1 reaches a steady state after a certain amount of time has passed since the start of operation, based on the rise characteristics of the tracking filter TFa. The torque vibration compensator 108a can correctly detect the magnitude of the torque vibration when the first tracking output signal Fout1 has risen to the steady state in this way.

[0114] 16, the monitoring circuit MNIa successively detects the amplitude of the first tracking output signal Fout1 and determines whether the difference between the previous amplitude A[t-1] and the current amplitude A[t] is smaller than a predetermined difference threshold ΔAth. If "|A[t]-A[t-1]|<ΔAth" is satisfied, the monitoring circuit MNIa determines that the first tracking output signal Fout1 is in a steady state.

[0115] 15A, when the monitoring circuit MNIa determines that the first tracking output signal Fout1 is in the steady state (step S103: Yes), it outputs a first learning start signal LST1 (step S104). Furthermore, the monitoring circuit MNIa sets the output amplitude of the first tracking output signal Fout1 at the time of determining that the steady state is in the steady state to the pre-suppression amplitude A[0] shown in FIG.

[0116] Meanwhile, the compensation value generation circuit 125-1 receives the first update amount UA1 via the selector switch SW2. In response to the first learning start signal LST1 in step S104, the compensation value generation circuit 125-1 starts a first learning operation based on the first update amount UA1 and therefore the first tracking output signal Fout1 (step S105). In conjunction with the first learning operation, the compensation value generation circuit 125-1 outputs the first torque compensation value TCV1 while sequentially updating it.

[0117] As a result, torque oscillation is gradually suppressed, and the monitoring circuit MNIa outputs a first learning completion signal LCP1 at some point in time. When the first learning completion signal LCP1 is output (step S106: Yes), the compensation value generation circuit 125-1 completes the first learning operation (step S107). Then, the compensation value generation circuit 125-1 continuously outputs the first torque compensation value TCV1 at the time the first learning operation is completed, i.e., the first torque compensation value TCV1 based on the first compensation value table CTBL1 at the time the first learning operation is completed.

[0118] Next, the torque vibration compensator 108a performs the process shown in Fig. 15B. In Fig. 15B, steps S201-S204 are processes for preparing for the second learning operation. Steps S205-S207 are processes accompanying the execution of the second learning operation. In the processes of steps S201-S207, the first-order tracking frequency ω r is the second-order tracking frequency (2ω r ) will be changed to

[0119] Briefly, in step S201, the torque vibration compensator 108a uses the changeover switches SW1 and SW2 to adjust the torque vibration to a double mechanical angle (2θm), that is, a second-order tracking frequency (2ω r ) is selected. Accordingly, the tracking filter TFa selects the second-order tracking frequency (2ω r ) to generate a second tracking output signal Fout2 (step S202).

[0120] Next, the torque vibration compensator 108a waits for the second tracking output signal Fout2 to reach a steady state (step S203). If the monitoring circuit MNIa determines that the second tracking output signal Fout2 is in a steady state (step S203: Yes), it outputs a second learning start signal LST2 (step S204). Furthermore, the monitoring circuit MNIa sets the output amplitude of the second tracking output signal Fout2 at the time when it determines that the second tracking output signal Fout2 is in a steady state to the pre-suppression amplitude A[0] shown in FIG. 11, for example.

[0121] Meanwhile, the compensation value generation circuit 125-2 receives the second update amount UA2 via the selector switch SW2. In response to the second learning start signal LST2 in step S204, the compensation value generation circuit 125-2 starts a second learning operation based on the second update amount UA2 and therefore the second tracking output signal Fout2 (step S205). In conjunction with the second learning operation, the compensation value generation circuit 125-2 outputs the second torque compensation value TCV2 while sequentially updating it.

[0122] As a result, torque oscillation is gradually suppressed, and the monitoring circuit MNIa outputs a second learning completion signal LCP2 at some point in time. When the second learning completion signal LCP2 is output (step S206: Yes), the compensation value generation circuit 125-2 completes the second learning operation (step S207). Then, the compensation value generation circuit 125-2 continuously outputs the second torque compensation value TCV2 at the time the second learning operation is completed, i.e., the second torque compensation value TCV2 based on the second compensation value table CTBL2 at the time the second learning operation is completed.

[0123] <About the torque vibration compensator (application example [2])> Fig. 17 is a schematic diagram showing an example of a configuration different from that of Fig. 14, which is an extension of the torque vibration compensator shown in Fig. 8. In Fig. 17, as in Fig. 14, a plurality of tracking frequencies, in this example, a first-order tracking frequency ω r and the second-order tracking frequency (2ω r) are provided. Accordingly, the torque vibration compensator 108b shown in Fig. 17, unlike the case of Fig. 14, is provided with two tracking filters TFa1 and TFa2 corresponding to the two tracking frequencies, respectively. Furthermore, the torque vibration compensator 108b is provided with two monitoring circuits MNIa1 and MNIa2 and two multipliers 124-1 and 124-2 corresponding to the two tracking filters TFa1 and TFa2, respectively.

[0124] The first tracking filter TFa1 receives the mechanical angle θm and generates a first-order tracking frequency ω r The first tracking output signal Fout1 is generated using the first monitoring circuit MNIa1. The first monitoring circuit MNIa1 generates a first learning completion signal LCP1 by monitoring the processing state of the first tracking filter TFa1. The multiplier 124-1 receives the first tracking output signal Fout1 and generates a first update amount UA1.

[0125] On the other hand, the second tracking filter TFa2 receives a doubled mechanical angle (2θm) and generates a second-order tracking frequency (2ω r ) to generate a second tracking output signal Fout2. A second monitoring circuit MNIa2 generates a second learning completion signal LCP2 by monitoring the processing state of the second tracking filter TFa2. A multiplier 124-2 receives the second tracking output signal Fout2 and generates a second update amount UA2.

[0126] 14, processing is then performed using the two compensation value generation circuits 125-1, 125-2, etc. Briefly, the compensation value generation circuit 125-1 starts a first learning operation using the first update amount UA1 and, in turn, the first tracking output signal Fout1 as input. Thereafter, the compensation value generation circuit 125-1 completes the first learning operation in response to a first learning completion signal LCP1, and continuously reflects the first torque compensation value TCV1 at the time of completion of the learning operation in the motor control signal.

[0127] Meanwhile, in parallel with the compensation value generation circuit 125-1, the compensation value generation circuit 125-2 starts a second learning operation using the second update amount UA2 and therefore the second tracking output signal Fout2 as input. Thereafter, the compensation value generation circuit 125-2 completes the second learning operation in response to the second learning completion signal LCP2, and continuously reflects the second torque compensation value TCV2 at the time of completion of the learning operation in the motor control signal.

[0128] By using the above-described configuration example, it is possible to suppress the first and second frequency components contained in torque vibration, for example, as in the case of FIG. 14. In the configuration example shown in FIG. 17, unlike the case of FIG. 14, the first and second frequency components are suppressed by parallel processing using separate resources, which can shorten the time required to suppress torque vibration, for example. On the other hand, in the configuration example shown in FIG. 14, unlike the case of FIG. 17, the first and second frequency components are suppressed by serial processing using shared resources, which can reduce the overhead of the circuit area or the processing load on the processor PRC. Note that in FIGS. 14 and 17, frequency components up to the second order are suppressed, but it is also possible to suppress third and subsequent frequency components in a similar manner.

[0129] Fig. 18A is a flow chart showing an example of the processing contents of the main parts of the torque vibration compensator 108b shown in Fig. 17. Fig. 18B is a flow chart showing an example of the processing contents subsequent to Fig. 18A. For example, the processor PRC in Fig. 5 executes these processes based on a program in the memory MEM. In the flow shown in Fig. 18A, the processing of step S101 is deleted from the flow shown in Fig. 15A. Furthermore, the processing of steps S102-S104 is performed by the first tracking filter TFa1 and the first monitoring circuit MNIa1.

[0130] Similarly, in the flow shown in Fig. 18B, the processing of step S201 is deleted from the flow shown in Fig. 15B. Also, unlike the case of Fig. 15B, the flow shown in Fig. 18B is executed in parallel with the flow shown in Fig. 18A. In Fig. 18B, the processing of steps S202-S204 is performed by the second tracking filter TFa2 and the second monitoring circuit MNIa2.

[0131] <About the torque vibration compensator (variant)> In the explanation so far, the torque vibration compensator 108 performs a learning operation to suppress torque vibration, for example, during the startup phase of the motor system, and after completing the learning operation, continuously outputs the torque compensation value TCV at the time of completion. In this case, the tracking filter TFa and the monitoring circuit MNIa do not need to operate particularly after the learning operation is completed. However, the tracking filter TFa and the monitoring circuit MNIa may continue to operate after the learning operation is completed. In this case, the torque vibration compensator 108 can perform the learning operation again, for example, when there is a significant change in the abnormality indexes IDXx, IDXy or the suppression rate SR.

[0132] <Examples of application to other systems> The semiconductor device 10 shown in FIG. 1 is not limited to applications such as torque vibration suppression in motor systems as described above, but can be used for a variety of purposes in a variety of systems. As an example, the semiconductor device 10 can be used to prevent malfunctions in a magnetic bearing control system. In this case, the tracking filter is mainly used to control the magnetic bearing. In controlling a magnetic bearing, a phenomenon (unbalanced force) occurs in which the shaft vibrates outward from its center position due to centrifugal force during high-speed rotation, which is mainly caused by a minute unbalanced weight that exists in the magnetically levitated shaft / rotor.

[0133] Here, the semiconductor device 10 can be applied to a compensation device for canceling the unbalance force. In this case, the semiconductor device 10 detects a tracking frequency ω rIn addition, when the AC component is large, the semiconductor device 10 can prevent overcompensation by disabling the function of canceling the centrifugal force.

[0134] <Major effects of each embodiment> As described above, the semiconductor device according to the embodiment includes a tracking filter including a low-pass filter and a monitoring circuit that determines whether or not there is an abnormality in the tracking filter by detecting an AC component included in the output signal of the low-pass filter. This makes it possible to prevent malfunctions in a system equipped with a tracking filter. This further enhances the safety of the system. In particular, by using the semiconductor device to suppress torque vibrations in a motor system, it is possible to prevent the tracking filter itself from functioning as an unintended vibration source. As a result, torque vibrations can be more reliably suppressed.

[0135] The invention made by the inventor has been specifically described above based on the embodiments, but the present invention is not limited to the above embodiments and can be modified in various ways without departing from the spirit of the invention. For example, the above-described embodiments have been described in detail to clearly explain the present invention, and the present invention is not necessarily limited to those including all of the described configurations. Furthermore, it is possible to replace part of the configuration of one embodiment with the configuration of another embodiment, or to add the configuration of another embodiment to the configuration of one embodiment. Furthermore, it is possible to add, delete, or replace part of the configuration of each embodiment with other configurations.

[0136] Furthermore, each unit is typically implemented by program processing using a CPU (Central Processing Unit). That is, each unit is implemented on the CPU by the CPU executing a program stored in memory. However, the implementation form of each unit is not limited to this software, and may be hardware such as an FPGA (Field Programmable Gate Array) or an ASIC (Application Specific Integrated Circuit), or may be a combination of software and hardware.

[0137] The above-mentioned program may be stored in a non-transitory, tangible, computer-readable recording medium and then supplied to a computer. Examples of such a recording medium include magnetic recording media such as hard disk drives, optical recording media such as DVDs (Digital Versatile Discs) and Blu-ray Discs, and semiconductor memories such as flash memories and SSDs (Solid State Drives). [Explanation of symbols]

[0138] 10 Semiconductor device 20 Inverter 102 Speed ​​Controller 108 Torque Vibration Compensator 125 Compensation value generation circuit ACx,ACy AC component DCx,DCy DC component FLG Abnormal flag signal FTH threshold Fin Tracking input signal Fout Tracking output signal Gpwm PWM signal (motor control signal) IDXx,IDXy Anomaly index ITHx, ITHy threshold LCP learning completion signal LPFx,LPFy Low-pass filters ML1x, ML1y, ML2x, ML2y multipliers MNI monitoring circuit MT motor PRC Processor SC suppression completion signal TCV Torque Compensation Value TF Tracking Filter ω rotation speed ω * Speed ​​command value ω r Tracking Frequency

Claims

1. a tracking filter that extracts a frequency component of a tracking frequency from a tracking input signal containing a plurality of frequency components, and generates a tracking output signal composed of the extracted frequency component of the tracking frequency; a monitoring circuit for monitoring the processing state of the tracking filter; and The tracking filter an extraction multiplier for multiplying the tracking input signal by a cosine wave signal or a sine wave signal of the tracking frequency; a low-pass filter that removes frequency components higher than a cutoff frequency from the output signal of the extraction multiplier; a restoration multiplier that generates the tracking output signal by multiplying the output signal of the low-pass filter by a cosine wave signal or a sine wave signal of the tracking frequency; Equipped with the monitoring circuit detects an AC component included in the output signal of the low-pass filter, and determines whether or not there is an abnormality in the tracking filter based on the magnitude of the detected AC component. Semiconductor device.

2. 2. The semiconductor device according to claim 1, The monitoring circuit detecting a DC component contained in the output signal of the low-pass filter; calculating a ratio of the AC component to the DC component as an anomaly index; determining whether or not the tracking filter is abnormal based on the abnormality index; Semiconductor device.

3. 3. The semiconductor device according to claim 2, The monitoring circuit determines that an abnormality exists when the abnormality index is higher than a predetermined threshold value, and determines that no abnormality exists when the abnormality index is lower than the threshold value. Semiconductor device.

4. 2. The semiconductor device according to claim 1, the extraction multiplier comprises a first multiplier that multiplies the cosine wave signal and a second multiplier that multiplies the sine wave signal; the low-pass filter comprises a first low-pass filter to which the output signal of the first multiplier is input, and a second low-pass filter to which the output signal of the second multiplier is input; the monitoring circuit detects a first AC component included in the output signal of the first low-pass filter and a second AC component included in the output signal of the second low-pass filter, and determines whether or not there is an abnormality in the tracking filter based on the magnitude of at least one of the first AC component and the second AC component. Semiconductor device.

5. 2. The semiconductor device according to claim 1, A plurality of said tracking frequencies are provided, the tracking filter generates the plurality of tracking output signals in a time division manner by switching the plurality of tracking frequencies. Semiconductor device.

6. 2. The semiconductor device according to claim 1, A plurality of said tracking frequencies are provided, The semiconductor device includes: a plurality of the tracking filters respectively corresponding to the plurality of tracking frequencies; a plurality of the monitoring circuits respectively corresponding to the plurality of tracking filters; Equipped with Semiconductor device.

7. A memory that stores a program; a processor that executes the program stored in the memory; Equipped with The processor, based on the program, (a) a tracking process for extracting a frequency component of a tracking frequency from a tracking input signal containing a plurality of frequency components, and generating a tracking output signal composed of the extracted frequency component of the tracking frequency; (b) a monitoring process for monitoring the processing status of the tracking process; Run In the tracking process of (a), (a1) multiplying the tracking input signal by a sine wave signal or a cosine wave signal at the tracking frequency; (a2) removing frequency components higher than a cutoff frequency from the output signal obtained in (a1); (a3) multiplying the output signal obtained in (a2) by a sine wave signal or a cosine wave signal of the tracking frequency to generate the tracking output signal; During the monitoring process of (b), an AC component included in the output signal obtained in (a2) is detected, and whether or not there is an abnormality in the tracking process is determined based on the magnitude of the detected AC component. Semiconductor device.

8. 8. The semiconductor device according to claim 7, The processor, during the monitoring process of (b), Detecting a DC component contained in the output signal obtained in (a2), calculating a ratio of the AC component to the DC component as an anomaly index; determining whether or not there is an abnormality in the tracking process based on the abnormality index; Semiconductor device.

9. 9. The semiconductor device according to claim 8, The processor determines that an abnormality exists when the abnormality index is higher than a predetermined threshold, and determines that an abnormality does not exist when the abnormality index is lower than the threshold. Semiconductor device.

10. A semiconductor device that outputs a motor control signal to an inverter that supplies power to a motor and controls the motor via the inverter, a tracking filter that receives a speed deviation between a preset speed command value and a value of the rotational speed of the motor as a tracking input signal, extracts a frequency component of a tracking frequency that is a target for torque vibration suppression from the tracking input signal, and generates a tracking output signal that is composed of the extracted frequency component of the tracking frequency; a monitoring circuit for monitoring the processing state of the tracking filter; a compensation value generation circuit that generates and successively updates a compensation value for suppressing the torque vibration through a learning operation using the tracking output signal as an input, and reflects the generated compensation value in the motor control signal; Equipped with The tracking filter an extraction multiplier for multiplying the tracking input signal by a sine wave signal or a cosine wave signal of the tracking frequency; a low-pass filter that removes frequency components higher than a cutoff frequency from the output signal of the extraction multiplier; a restoration multiplier that generates the tracking output signal by multiplying the output signal of the low-pass filter by a sine wave signal or a cosine wave signal of the tracking frequency; and the monitoring circuit detects an AC component included in the output signal of the low-pass filter, determines whether or not there is an abnormality in the tracking filter based on the magnitude of the detected AC component, and generates a learning completion signal when it determines that there is an abnormality; the compensation value generation circuit completes the learning operation in response to the learning completion signal. Semiconductor device.

11. 11. The semiconductor device according to claim 10, the monitoring circuit further detects the amplitude of the tracking output signal, determines whether the suppression of the torque vibration has been completed or not based on the magnitude of the detected output amplitude, and generates the learning completion signal even when it determines that the suppression has been completed; the compensation value generating circuit completes the learning operation in response to the learning completion signal, and continuously reflects the compensation value at the time of completing the learning operation in the motor control signal. Semiconductor device.

12. 12. The semiconductor device according to claim 11, The monitoring circuit further comprises: detecting a DC component contained in the output signal of the low-pass filter; calculating a ratio of the AC component to the DC component as an anomaly index; When the abnormality index is higher than a predetermined first threshold, it is determined that an abnormality exists, and when the abnormality index is lower than the first threshold, it is determined that an abnormality does not exist. Semiconductor device.

13. 12. The semiconductor device according to claim 11, The monitoring circuit defines the magnitude of the output amplitude at the start of the learning operation as a pre-suppression amplitude, and while sequentially detecting the output amplitude during the learning operation, sequentially calculates a ratio of the detected output amplitude to the pre-suppression amplitude as a suppression rate, and determines that suppression is complete when the suppression rate is lower than a predetermined second threshold, and determines that suppression is incomplete when the suppression rate is higher than the second threshold. Semiconductor device.

14. 13. The semiconductor device according to claim 12, a plurality of said tracking frequencies are provided, including a first tracking frequency and a second tracking frequency; the tracking filter generates a first tracking output signal using the first tracking frequency, and generates a second tracking output signal using the second tracking frequency in response to a first learning completion signal that is one of the learning completion signals; The compensation value generation circuit starting a first learning operation using the first tracking output signal as an input; completing the first learning operation in response to the first learning completion signal, continuously reflecting the first compensation value at the time of completing the first learning operation in the motor control signal, and starting a second learning operation using the second tracking output signal as an input; completing the second learning operation in response to a second learning completion signal, which is another of the learning completion signals, and continuously reflecting a second compensation value at the time when the second learning operation is completed in the motor control signal; Semiconductor device.

15. 15. The semiconductor device according to claim 14, the second tracking frequency is twice the frequency of the first tracking frequency; Semiconductor device.

16. 13. The semiconductor device according to claim 12, a plurality of said tracking frequencies are provided, including a first tracking frequency and a second tracking frequency; The semiconductor device includes: a first tracking filter and a second tracking filter corresponding to the first tracking frequency and the second tracking frequency, respectively; a first monitoring circuit and a second monitoring circuit corresponding to the first tracking filter and the second tracking filter, respectively; Equipped with The compensation value generation circuit a first learning operation is started using a first tracking output signal from the first tracking filter as an input, the first learning operation is completed in response to a first learning completion signal from the first monitoring circuit, and a first compensation value at the time of completion of the first learning operation is continuously reflected in the motor control signal; a second learning operation is started using a second tracking output signal from the second tracking filter as an input, the second learning operation is completed in response to a second learning completion signal from the second monitoring circuit, and a second compensation value at the time of completion of the second learning operation is continuously reflected in the motor control signal; Semiconductor device.

17. 17. The semiconductor device according to claim 16, the second tracking frequency is twice the frequency of the first tracking frequency; Semiconductor device.

18. 11. The semiconductor device according to claim 10, a speed controller that generates a torque command value so that a speed deviation between the speed command value and a value of the rotational speed of the motor approaches zero; the compensation value generation circuit adds the generated compensation value to the torque command value. Semiconductor device.

Images