Switch noise suppression method for motor controller

Abstract

The invention discloses a motor controller switching noise suppression method, which comprises the following steps of: executing a task I and a task II in parallel after an FOC interrupt function is triggered by PWM (Pulse Width Modulation); task 1: performing a traditional FOC algorithm, and adjusting a trigger mode of ADC current sampling at a software bottom layer to be a PWM period zero point and midpoint dual trigger mode to realize dual sampling and dual updating of three-phase current, and generating PWM modulation waves; task 2: generating a random triangular carrier frequency, and then performing Markov chain-based two-state probability number arrangement on random numbers uniformly distributed in the triangular carrier frequency to obtain a random number group obeying a two-state Markov chain; and generating a PWM switch control signal through the PWM modulation wave obtained by the task 1 and the random number group obeying the two-state Markov chain obtained by the task 2. According to the invention, the noise level of the switching frequency and the integer frequency multiplication of the switching frequency can be effectively reduced, and the problem of overlarge output torque fluctuation can be reduced.

Description

Technical Field

[0001] The invention relates to a noise suppression method, in particular to a method for suppressing switch noise of a motor controller, and belongs to the technical field of motor control. Background Art

[0002] As electrification is applied in more and more areas of people's lives, its inherent high-frequency noise problem also needs to be solved urgently. In the application field of motor variable frequency power supply, due to the widespread application of pulse width modulation PWM technology, a large number of current harmonics will be introduced into the armature winding of the motor, resulting in obvious high-frequency noise near the switching frequency and its multiples of the motor. For the new energy EV industry, the switching frequency of the main drive motor controller for passenger cars is generally set to 10khz or nearby. By analyzing the noise spectrum when the controller is working in the noise laboratory, it can be obtained that there is obvious noise in all speed ranges near the frequency of 10khz. The noise shows the characteristics of umbrella-shaped distribution on the spectrum diagram. Moreover, during the operation of the controller, through the Fourier analysis of the phase current of the built-in permanent magnet synchronous motor, it can also be observed on the oscilloscope that there are obvious amplitude spikes at the switching frequency and its multiples of the controller.

[0003] There are many methods to suppress the switching noise, among which the most commonly used method is the random PWM method. Specifically, it includes random carrier, random PWM pulse position and random PWM frequency methods. It also includes some novel composite methods, such as double random PWM, that is, the method of random PWM pulse position superimposed on random PWM frequency. It also includes improved methods, such as the zero vector-variable delay random PWM method and research on random numbers, including the use of uniform random, Gaussian normal distribution random and other methods.

[0004] For example, “Research on Dual Random PWM Technology Based on Vector Control System” (see “Proceedings of the CSEE”, Vol. 30, No. 36, 2010) proposed a zero vector-variable delay random PWM method with a fixed sampling period, which can evenly distribute the current spectrum over a wider range, reduce the harmonic amplitude of integer multiple switching frequencies, effectively reduce system noise and suppress electromagnetic interference. However, the random effect of this random method deteriorates when the motor modulation ratio is high, and it only has an inhibitory effect on the high-frequency vibration of the motor.

[0005] For example, Patent Publication No. CN114844440 of "A Method for Eliminating Electromagnetic Noise at a Specific Frequency of a Motor" proposes a method for suppressing switching noise. It mainly randomizes the switch position of PWM by calculating the relationship between the switch position and the duty cycle during PWM modulation, thereby suppressing switching noise. This method is essentially still a random strategy for pulse position in random PWM, but it has a large amount of calculation, which poses a great challenge to the software computing load consumption, especially when the switching frequency is high, it will occupy more computing resources.

[0006] It can be seen from this that the existing motor noise suppression method has poor random effect and a narrow frequency adaptation range, and has a large amount of calculation and occupies a lot of computing resources. Summary of the invention

[0007] The technical problem to be solved by the present invention is to provide a method for suppressing switching noise of a motor controller, which can effectively reduce the noise video and solve the problem of excessive output torque fluctuation.

[0008] In order to solve the above technical problems, the technical solution adopted by the present invention is: A method for suppressing switching noise of a motor controller comprises the following steps: S1, after PWM triggers the FOC interrupt function, execute Task 1 and Task 2 in parallel; Task 1: Perform the traditional FOC algorithm and adjust the ADC current sampling trigger mode to the PWM cycle zero point and midpoint dual trigger mode to achieve dual sampling and dual update of the three-phase current and generate PWM modulation wave. Task 2: Generate a random triangular carrier frequency, and then sort the uniformly distributed random numbers in the triangular carrier frequency by the two-state probability number based on the Markov chain to obtain a random array that obeys the two-state Markov chain; S2. Generate a PWM switch control signal through the PWM modulation wave obtained in task one and the random array obeying the two-state Markov chain obtained in task two.

[0009] Furthermore, the traditional FOC algorithm in step S1 includes current sampling, voltage sampling, rotor angle sampling, coordinate system transformation, PI regulation and SVPWM space vector modulation algorithm.

[0010] Furthermore, task one in step S1 is specifically: 1.1. The PWM channel triggers the FOC interrupt function and the ADC current sampling of the three-phase current; 1.2. Set the carrier mode to Up-Down mode; 1.3. Set the ADC sampling trigger mode to perform ADC sampling trigger at the zero point and the maximum point of the PWM counting cycle respectively; 1.4. Set the FOC interrupt trigger mode to trigger the FOC interrupt at the zero point and maximum point of the PWM counting cycle respectively.

[0011] Furthermore, the random triangular carrier frequency is generated in step S1 by generating a random number array G[] in the range of -1 to 1 that obeys a uniform distribution through a uniform random number generator, and then multiplying the random number array G[] that obeys a uniform distribution by a random frequency gain value k, and then adding the minimum value f of the frequency range. min , thus obtaining a random number f that obeys a uniform distribution s , ; in, , f min is the minimum value of the frequency range, f max is the maximum value of the frequency range, G[i] is a random number in the random number array G[] that obeys uniform distribution; A random number f that follows a uniform distribution s The value range is F s ±F s / 4,F s is the original fixed center switching frequency.

[0012] Furthermore, in step S1, the random numbers uniformly distributed in the triangular carrier frequency are sorted by the two-state probability number based on the Markov chain to obtain a random number array that obeys the two-state Markov chain: 2.1. Assume that R[i]>0 is S1 state, R[i]<0 is S2 state; R[i] is a random number f that follows uniform distribution s Random numbers in ; 2.2. For any random number f from a uniform distribution s The random number R[i] (i=0, 1, 2, ...) is determined to determine whether R[i] is greater than the preset threshold m. , P is the transition probability, and ; 2.3. If R[i] is greater than the preset threshold m, and the previous random number R[i-1] of the random number R[i] is in the S1 state, the current random number M[i] after the transfer is in the S2 state, that is, M[i]=-R[i]; 2.4. If R[i] is greater than the preset threshold m, and the previous random number R[i-1] of the random number R[i] is in the S1 state, the current random number M[i] after the transfer is in the S1 state, that is, M[i]=R[i]; 2.5. The transferred random array M[] is multiplied by the gain G to obtain the random array GMarkov[], that is, GMarkov[]=M[]*G; 2.6. The final random array is Rfinal[], ; Among them, F s is the original fixed center switching frequency.

[0013] Furthermore, step S2 is specifically as follows: the frequency of the random array that obeys the two-state Markov chain obtained through task two, that is, the final random array Rfinal[], is a finite random change, and then the frequency of the final random array Rfinal[] is converted into a cycle count value that conforms to PWM, and a random number frequency-PWM cycle index comparison table is established for any random number frequency Rfinal[i] in the final random array Rfinal[] and the cycle count value that conforms to PWM that corresponds one to one with Rfinal[i], and the next PWM switching cycle value is obtained in real time by repeatedly looking up the table.

[0014] Furthermore, the step S2 is specifically as follows: dividing the zero vector of the PWM modulation wave obtained in task 1 into two even segments and setting them into three states: The first state: Both zero vectors appear at the starting point of the PWM modulation wave; The second state: Both zero vectors appear at the end point of the PWM modulation wave; The third state: a zero vector appears at the starting point of the PWM modulation wave, and another zero vector appears at the ending point of the PWM modulation wave; The random array obeying the two-state Markov chain obtained in Task 2 is matched to the three states of the zero vector of the PWM modulation wave to generate a PWM switch control signal whose random method is a Markov chain.

[0015] Compared with the prior art, the present invention has the following advantages and effects: the present invention provides a method for suppressing switching noise of a motor controller, which samples the PWM frequency or pulse position random strategy based on a Markov chain, which can not only effectively reduce the noise level of the switching frequency and its integer multiples, but also help to reduce the problem of excessive output torque fluctuation; at the same time, by adjusting the control density of the random PWM and increasing the control frequency of the PWM, specifically, by adopting a double sampling and double updating method of the motor phase current, the distortion rate THD of the motor output phase current and the output torque fluctuation are reduced, thereby further reducing the noise vibration level and torque fluctuation, and solving the problem of increased output torque fluctuation caused by the use of a random PWM algorithm alone. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 The present invention is a flow chart of a method for suppressing switching noise of a motor controller.

[0017] Figure 2 It is a frequency distribution diagram of random numbers that obey uniform distribution of the present invention.

[0018] Figure 3 It is a frequency distribution diagram of a random array obeying a two-state Markov chain of the present invention.

[0019] Figure 4 This is a schematic diagram of three states of the zero vector in Example 2 of a method for suppressing switching noise of a motor controller of the present invention.

[0020] Figure 5 This is the FFT result diagram of the original algorithm's phase current Fourier analysis.

[0021] Figure 6 It is a diagram of the phase current Fourier analysis FFT result of the present invention.

[0022] Figure 7 Schematic diagram of the influence of different control strategies on torque fluctuation in an embodiment of the present invention.

[0023] Figure 8 Schematic diagram of the effect of different control strategies on phase current THD in an embodiment of the present invention. DETAILED DESCRIPTION

[0024] In order to elaborate on the technical scheme adopted by the present invention to achieve the predetermined technical purpose, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only partial embodiments of the present invention, rather than all embodiments, and the technical means or technical features in the embodiments of the present invention can be replaced without paying creative work. The present invention will be described in detail below with reference to the drawings and in conjunction with the embodiments.

[0025] Example 1: Figure 1 As shown, a method for suppressing switching noise of a motor controller of the present invention comprises the following steps: S1. After PWM triggers the FOC interrupt function, Task 1 and Task 2 are executed in parallel.

[0026] Task 1: Perform the traditional FOC algorithm, and adjust the ADC current sampling trigger mode to the PWM cycle zero point and midpoint dual trigger mode to achieve double sampling and double update of the three-phase current to generate a PWM modulation wave.

[0027] The traditional FOC algorithm in step S1 includes current sampling, voltage sampling, rotor angle sampling, coordinate system transformation, PI regulation and SVPWM space vector modulation algorithm.

[0028] Task 1 in step S1 is specifically: 1.1. The PWM channel triggers the FOC interrupt function and the ADC current sampling of the three-phase current; 1.2. Set the carrier mode to Up-Down mode; 1.3. Set the ADC sampling trigger mode to perform ADC sampling trigger at the zero point and the maximum point of the PWM counting cycle respectively; 1.4. Set the FOC interrupt trigger mode to trigger the FOC interrupt at the zero point and maximum point of the PWM counting cycle respectively.

[0029] The random triangular carrier frequency is generated in step S1 as follows: a uniform random number generator is used to generate a random number array G[] in the range of -1 to 1 that obeys a uniform distribution, and then the random number array G[] that obeys a uniform distribution is multiplied by the random frequency gain value k, and then the minimum value f of the frequency range is added. min , thus obtaining Figure 2 The random number f shown is uniformly distributed s , ; in, , f min is the minimum value of the frequency range, f max is the maximum value of the frequency range, G[i] is a random number in the random number array G[] that obeys uniform distribution; A random number f that follows a uniform distribution s The value range is F s ±F s / 4,F s is the original fixed center switching frequency.

[0030] In step S1, the random numbers uniformly distributed in the triangular carrier frequency are sorted according to the two-state probability number of the Markov chain to obtain a random array that obeys the two-state Markov chain: 2.1. Assume that R[i]>0 is S1 state, R[i]<0 is S2 state; R[i] is a random number f that follows uniform distribution s Random numbers in ; 2.2. For any random number f from a uniform distribution s The random number R[i] (i=0, 1, 2, ...) is determined to determine whether R[i] is greater than the preset threshold m. , P is the transition probability, and ; 2.3. If R[i] is greater than the preset threshold m, and the previous random number R[i-1] of the random number R[i] is in the S1 state, the current random number M[i] after the transfer is in the S2 state, that is, M[i]=-R[i]; 2.4. If R[i] is greater than the preset threshold m, and the previous random number R[i-1] of the random number R[i] is in the S1 state, the current random number M[i] after the transfer is in the S1 state, that is, M[i]=R[i]; 2.5. The transferred random array M[] is multiplied by the gain G to obtain the random array GMarkov[], that is, GMarkov[]=M[]*G; 2.6, such as Figure 3 As shown, the final random array is Rfinal[]. ; Among them, F s is the original fixed center switching frequency.

[0031] Task 2: Generate a random triangular carrier frequency, and then sort the uniformly distributed random numbers in the triangular carrier frequency using the two-state probability number of the Markov chain to obtain a random array that obeys the two-state Markov chain.

[0032] S2. Generate a PWM switch control signal through the PWM modulation wave obtained in task one and the random array obeying the two-state Markov chain obtained in task two.

[0033] Step S2 is specifically as follows: the frequency of the random array that obeys the two-state Markov chain obtained through Task 2, that is, the final random array Rfinal[], is a finite random change, and then the frequency of the final random array Rfinal[] is converted into a cycle count value that conforms to PWM, and a random number frequency-PWM cycle index comparison table is established for any random number frequency Rfinal[i] in the final random array Rfinal[] and the cycle count value that conforms to PWM that corresponds one to one with Rfinal[i], and the next PWM switching cycle value is obtained in real time by repeatedly looking up the table.

[0034] Example 2: Figure 1 As shown, a method for suppressing switching noise of a motor controller of the present invention comprises the following steps: S1. After PWM triggers the FOC interrupt function, Task 1 and Task 2 are executed in parallel.

[0035] Task 1: Perform the traditional FOC algorithm, and adjust the ADC current sampling trigger mode to the PWM cycle zero point and midpoint dual trigger mode to achieve double sampling and double update of the three-phase current to generate a PWM modulation wave.

[0036] The traditional FOC algorithm in step S1 includes current sampling, voltage sampling, rotor angle sampling, coordinate system transformation, PI regulation and SVPWM space vector modulation algorithm.

[0037] Task 1 in step S1 is specifically: 1.1. The PWM channel triggers the FOC interrupt function and the ADC current sampling of the three-phase current; 1.2. Set the carrier mode to Up-Down mode; 1.3. Set the ADC sampling trigger mode to perform ADC sampling trigger at the zero point and the maximum point of the PWM counting cycle respectively; 1.4. Set the FOC interrupt trigger mode to trigger the FOC interrupt at the zero point and maximum point of the PWM counting cycle respectively.

[0038] The random triangular carrier frequency is generated in step S1 as follows: a uniform random number generator is used to generate a random number array G[] in the range of -1 to 1 that obeys a uniform distribution, and then the random number array G[] that obeys a uniform distribution is multiplied by the random frequency gain value k, and then the minimum value f of the frequency range is added. min , thus obtaining Figure 2 The random number f shown is uniformly distributed s , ; in, , f min is the minimum value of the frequency range, f max is the maximum value of the frequency range, G[i] is a random number in the random number array G[] that obeys uniform distribution; A random number f that follows a uniform distribution s The value range is F s ±F s / 4,F s is the original fixed center switching frequency.

[0039] In step S1, the random numbers uniformly distributed in the triangular carrier frequency are sorted according to the two-state probability number of the Markov chain to obtain a random array that obeys the two-state Markov chain: 2.1. Assume that R[i]>0 is S1 state, R[i]<0 is S2 state; R[i] is a random number f that follows uniform distribution s Random numbers in ; 2.2. For any random number f from a uniform distribution s The random number R[i] (i=0, 1, 2, ...) is determined to determine whether R[i] is greater than the preset threshold m. , P is the transition probability, and ; 2.3. If R[i] is greater than the preset threshold m, and the previous random number R[i-1] of the random number R[i] is in the S1 state, the current random number M[i] after the transfer is in the S2 state, that is, M[i]=-R[i]; 2.4. If R[i] is greater than the preset threshold m, and the previous random number R[i-1] of the random number R[i] is in the S1 state, the current random number M[i] after the transfer is in the S1 state, that is, M[i]=R[i]; 2.5. The transferred random array M[] is multiplied by the gain G to obtain the random array GMarkov[], that is, GMarkov[]=M[]*G; 2.6, such as Figure 3 As shown, the final random array is Rfinal[]. ; Among them, F s is the original fixed center switching frequency.

[0040] Task 2: Generate a random triangular carrier frequency, and then sort the uniformly distributed random numbers in the triangular carrier frequency using the two-state probability number of the Markov chain to obtain a random array that obeys the two-state Markov chain.

[0041] S2. Generate a PWM switch control signal through the PWM modulation wave obtained in task one and the random array obeying the two-state Markov chain obtained in task two.

[0042] Step S2 is specifically as follows: Figure 4 As shown, the zero vector of the PWM modulation wave obtained in Task 1 is divided into two even segments and set to three states: The first state: Both zero vectors appear at the starting point of the PWM modulation wave; The second state: Both zero vectors appear at the end point of the PWM modulation wave; The third state: a zero vector appears at the starting point of the PWM modulation wave, and another zero vector appears at the ending point of the PWM modulation wave; The random array obeying the two-state Markov chain obtained in Task 2 is matched to the three states of the zero vector of the PWM modulation wave to generate a PWM switch control signal whose random method is a Markov chain.

[0043] In the prior art, the PWM control signal is compared with a triangular carrier by a modulation signal (the modulation signal is a sine wave, but because the actual comparison time accounts for a very small proportion of the period of the sine wave, it is regarded as a straight line at the microscopic level). When the amplitude of the triangular carrier is greater than the modulation signal, the PWM control signal is 0 (i.e., zero vector), and when the amplitude of the triangular carrier is less than the modulation signal, the PWM control signal is 1. The frequency of the traditional triangular carrier is fixed, so the zero vector appears fixedly at the starting point and the end point of the PWM modulation wave. The present invention randomizes the triangular carrier through a 2-state Markov chain, so that the frequency of the triangular carrier changes in a 2-state Markov chain, thereby causing the zero vector of the final PWM modulation wave to shift forward or backward, thereby forming the three zero vector distribution states of the present invention.

[0044] For the above-mentioned Embodiment 1 and Embodiment 2, FFT analysis is performed on the phase current and compared with the original algorithm of the traditional uniform random PWM control.

[0045] like Figure 5 As shown in the figure, the original algorithm: the motor phase current has a PSD peak of -26.58db near 10khz and a PSD peak of -10.86db near 20khz. Figure 6 As shown, according to the present invention: the PSD near 10 khz is reduced to -31.09 db, and the PSD near 20 khz is reduced to -20.02 db.

[0046] In summary, the present invention reduces the PSD near 10 kHz by 4.5 db, and the PSD near 20 kHz by 10 db. Moreover, the noise PSD near even multiples of the switch is also significantly suppressed, with significant effects.

[0047] like Figure 7 and 8 As shown in the figure, the common single sampling and double sampling torque fluctuations are the smallest, but they cannot effectively improve the switching frequency noise. After adding the common random strategy, although the switching noise is suppressed, the output torque fluctuation becomes significantly larger, as shown in the figure. Figure 7 As shown in the 7th experiment in , the torque is 10Nm. After sampling the Markov and double sampling algorithms claimed in this patent, the torque will be 5Nm, as shown in Figure 7 As shown in the 3rd and 4th experiments.

[0048] Based on the above test results, it can be seen that the current double sampling combined with the Markov chain random PWM strategy has the best effect of suppressing torque fluctuations and improving phase current distortion rate. This method has the best practical effect in the control method of suppressing 10khz noise.

[0049] The above is only a preferred embodiment of the present invention and does not limit the present invention in any form. Although the present invention has been disclosed as a preferred embodiment as above, it is not used to limit the present invention. Any technician familiar with this profession can make some changes or modify the technical contents disclosed above into equivalent embodiments without departing from the scope of the technical solution of the present invention. However, any simple modification, equivalent replacement and improvement made to the above embodiments without departing from the content of the technical solution of the present invention, based on the technical essence of the present invention, within the spirit and principles of the present invention, still fall within the protection scope of the technical solution of the present invention.

Claims

1. A method for suppressing switching noise of a motor controller, characterized in that The following steps are involved: S1, after PWM triggers the FOC interrupt function, execute Task 1 and Task 2 in parallel; Task 1: Perform the traditional FOC algorithm and adjust the ADC current sampling trigger mode to the PWM cycle zero point and midpoint dual trigger mode to achieve dual sampling and dual update of the three-phase current and generate PWM modulation wave. Task 2: Generate a random triangular carrier frequency, and then sort the uniformly distributed random numbers in the triangular carrier frequency by the two-state probability number based on the Markov chain to obtain a random array that obeys the two-state Markov chain; S2. Generate a PWM switch control signal through the PWM modulation wave obtained in task one and the random array obeying the two-state Markov chain obtained in task two.

2. A method for suppressing switching noise of a motor controller according to claim 1, characterized in that: The traditional FOC algorithm in step S1 includes current sampling, voltage sampling, rotor angle sampling, coordinate system transformation, PI regulation and SVPWM space vector modulation algorithm.

3. A method for suppressing switching noise of a motor controller according to claim 1, characterized in that: Task 1 in step S1 is specifically: 1.

1. The PWM channel triggers the FOC interrupt function and the ADC current sampling of the three-phase current; 1.

2. Set the carrier mode to Up-Down mode; 1.

3. Set the ADC sampling trigger mode to perform ADC sampling trigger at the zero point and maximum point of the PWM counting cycle respectively; 1.

4. Set the FOC interrupt trigger mode to trigger the FOC interrupt at the zero point and maximum point of the PWM counting cycle respectively.

4. The method for suppressing switching noise of a motor controller according to claim 1, characterized in that: The step S1 of generating the random triangular carrier frequency is as follows: a uniform random number generator is used to generate a random number array G[] in the range of -1 to 1 that obeys a uniform distribution, and then the random number array G[] that obeys a uniform distribution is multiplied by a random frequency gain value k, and then the minimum value f of the frequency range is added. min , thus obtaining a random number f that obeys a uniform distribution s , ； in, , f min is the minimum value of the frequency range, f max is the maximum value of the frequency range, G[i] is a random number in the random number array G[] that obeys uniform distribution; A random number f that follows a uniform distribution s The value range is F s ±F s / 4,F s is the original fixed center switching frequency.

5. The method for suppressing switching noise of a motor controller according to claim 1, characterized in that: In step S1, the random numbers uniformly distributed in the triangular carrier frequency are sorted by the two-state probability numbers based on the Markov chain to obtain a random number array that obeys the two-state Markov chain: 2.

1. Assume that R[i]>0 is S1 state, R[i]<0 is S2 state; R[i] is a random number f that follows uniform distribution s Random numbers in ; 2.

2. For any random number f from a uniform distribution s The random number R[i] (i=0, 1, 2, ...) is determined to determine whether R[i] is greater than the preset threshold m. , P is the transition probability, and ; 2.

3. If R[i] is greater than the preset threshold m, and the previous random number R[i-1] of the random number R[i] is in the S1 state, the current random number M[i] after the transfer is in the S2 state, that is, M[i]=-R[i]; 2.

4. If R[i] is greater than the preset threshold m, and the previous random number R[i-1] of the random number R[i] is in the S1 state, the current random number M[i] after the transfer is in the S1 state, that is, M[i]=R[i]; 2.

5. The transferred random array M[] is multiplied by the gain G to obtain the random array GMarkov[], that is, GMarkov[]=M[]*G; 2.

6. The final random array is Rfinal[], ； Among them, F s is the original fixed center switching frequency.

6. The method for suppressing switching noise of a motor controller according to claim 1, characterized in that: The step S2 is specifically as follows: the frequency of the random array that obeys the two-state Markov chain obtained through task 2, that is, the final random array Rfinal[], is a finite random change, and then the frequency of the final random array Rfinal[] is converted into a cycle count value that conforms to PWM, and a random number frequency-PWM cycle index comparison table is established for any random number frequency Rfinal[i] in the final random array Rfinal[] and the cycle count value that conforms to PWM that corresponds one to one with Rfinal[i], and the next PWM switching cycle value is obtained in real time by repeatedly looking up the table.

7. The method for suppressing switching noise of a motor controller according to claim 1, characterized in that: The step S2 specifically includes: dividing the zero vector of the PWM modulation wave obtained in task 1 into two even segments and setting them into three states: The first state: Both zero vectors appear at the starting point of the PWM modulation wave; The second state: Both zero vectors appear at the end point of the PWM modulation wave; The third state: a zero vector appears at the starting point of the PWM modulation wave, and another zero vector appears at the ending point of the PWM modulation wave; The random array obeying the two-state Markov chain obtained in Task 2 is matched to the three states of the zero vector of the PWM modulation wave to generate a PWM switch control signal whose random method is a Markov chain.

Images