Self-adaptive controller for suppressing current harmonics

Abstract

The invention discloses a self-adaptive controller for suppressing current harmonics, which is characterized in that each harmonic component in d-axis current deviation deltaid and q-axis current deviation deltaiq is extracted, the amplitude of each injection harmonic is adjusted through a self-adaptive harmonic compensation module, and finally, the injection current of each harmonic compensation module is added and injected into the input of a current loop controller. Therefore, the purpose of minimizing each harmonic content in the d-axis current deviation delta id and the q-axis current deviation delta iq is achieved, and the corresponding harmonic component in the phase current can be suppressed. By suppressing harmonic waves in a d-axis current feedback value id and a q-axis current feedback value iq, the purpose of suppressing harmonic waves in a motor U-phase feedback current iu, a motor V-phase feedback current iv and a motor W-phase feedback current iw in three-phase current is achieved.

Description

Technical Field

[0001] The invention relates to a permanent magnet synchronous motor, and in particular to an adaptive controller for suppressing current harmonics used in the permanent magnet synchronous motor. Background Art

[0002] Permanent magnet synchronous motors are widely used in industrial fields such as robots, ships, and automobiles. Due to the nonlinear characteristics of the inverter and the presence of air gap magnetic field distortion, the three-phase voltage, current, and electromotive force waveforms of the motor will be distorted, thereby generating harmonic components of the phase voltage. The presence of harmonic voltage will cause the three-phase current in the motor to contain corresponding harmonic components, which are mainly reflected in the 5th, 7th, 11th, 13th, etc. (such as Figure 1 When the motor is running at a constant speed, the three-phase current can be expressed as follows. u 、i v 、i w are the three corresponding phase currents u, v, and w, w is the motor angular velocity, t is the current time, and I 1 is the fundamental current amplitude, I 5 ,I 7 are the 5th harmonic current amplitude and the 7th harmonic current amplitude respectively; θ 1 is the phase angle corresponding to the fundamental current at time 0, θ 5 is the phase angle corresponding to the 5th harmonic current at time 0, θ 7 It is the phase angle corresponding to the 7th harmonic current at time 0.

[0003]

[0004] According to the following formula, perform 3-2 transformation to convert the stationary three-phase coordinates into dq synchronous rotating coordinates:

[0005]

[0006] Simplifying, we get:

[0007]

[0008] It can be seen that the 5th and 7th harmonic components in the three-phase stationary coordinate system appear as the 6th harmonic in the dq synchronous rotating coordinate system. Similarly, the 11th and 13th harmonic components in the three-phase stationary coordinate system appear as the 12th harmonic in the dq synchronous rotating coordinate system. Most of the existing suppression algorithms operate synchronously with the control cycle, which requires more control computing power. In addition, the existing suppression algorithms are not effective when performing multiple harmonic suppression. Summary of the invention

[0009] The present invention aims to solve the problems that the suppression algorithm of the current harmonics in the existing permanent magnet synchronous motor needs to occupy a lot of control computing power; the effect is not good when multiple harmonics are suppressed, and a method for compensating the harmonics of the phase current in the dq coordinate system is provided. By suppressing the d-axis current feedback value i d and q-axis current feedback value i q Medium harmonics, to suppress the motor U-phase feedback current i in the three-phase current u , motor V phase feedback current i v and the motor W phase feedback current i w An adaptive controller is used to suppress current harmonics for the purpose of medium harmonics.

[0010] The present invention adopts an adaptive controller for suppressing current harmonics to solve the above technical problems, including a Park inverse converter, a Park converter, a controlled object P(Z), a d-axis proportional integral (PI) controller PI 1 , q-axis PI controller PI 2 , Clark converter, d-axis current deviation Δi d With the d-axis injection current i dinj The differential output d-axis PI controller input current i din To d-axis PI controller PI 1 , q-axis current deviation Δi q With the q-axis injection current i qinj The differential output q-axis PI controller input current i qin To q-axis PI controller PI 2 , P.I. 1 The output d-axis control voltage u d ,PI 2 The output q-axis control voltage u q ,d-axis control voltage u d and q-axis control voltage u q After the park inverse converter, the α-axis control voltage u is output α and β-axis control voltage u β , input the controlled object P(Z), the controlled object P(Z) feeds back the motor U phase feedback current i in the three-phase current of the motor u , motor V phase feedback current i v and motor W phase feedback current i w To the Clark converter, after being processed and transformed by the Clark converter, the α-axis feedback current i is output α and the β-axis feedback current i β To the park converter, after the park converter processing and transformation, the d-axis current feedback value i is output respectively. d and q-axis current feedback value i q , where the d-axis feedback current i dWith d-axis current given value The differential output d-axis current deviation Δi d , q-axis current feedback value i q With q-axis current given value The differential output q-axis current deviation Δi q ; d-axis current deviation Δi d and q-axis current deviation Δi q Input to the time domain → angular position domain mapper, after being processed by the time domain → angular position domain mapper, the d-axis current deviation with position domain coordinate n is output and the q-axis current deviation with position domain coordinate n Then it is input into the synchronous low-pass filter, and after filtering by the synchronous low-pass filter, the output is the d-axis current deviation I with the filtered position domain coordinate n. d (n) and the q-axis current deviation I with the filtered position domain coordinate n q (n) to k th Harmonic compensator, k th The harmonic compensator outputs the d-axis PI controller injection current i dinj and the q-axis PI controller injects current i qinj ; By suppressing the d-axis current feedback value i d and q-axis current feedback value i q Medium harmonics, to suppress the motor U-phase feedback current i in the three-phase current u , motor V phase feedback current i v and the motor W phase feedback current i w The purpose of harmonics.

[0011] The harmonics of the phase current can be compensated in the dq coordinate system by suppressing the d-axis current feedback value i d and q-axis current feedback value i q Medium harmonics, to suppress the motor U-phase feedback current i in the three-phase current u , motor V phase feedback current i v and the motor W phase feedback current i w The purpose of harmonics. Using the synchronous average filter in the position domain, the high-frequency noise can be filtered out while effectively extracting the harmonic components related to the motor running speed; using the uncoupled harmonic compensator, multiple harmonic components can be effectively compensated at the same time.

[0012] Preferably, the k th The harmonic compensator includes a first adaptive iterator corresponding to the d-axis and a second adaptive iterator corresponding to the q-axis, forming a d-axis k th Harmonic compensator and q-axis k th Harmonic compensator, where d axis k th Harmonic compensator and q-axis k thThe harmonic compensators all use the same corresponding algorithm implementation steps, where the q axis k th The harmonic compensator adopts the following algorithm implementation steps:

[0013] A1. Discretization of angular position domain;

[0014] The harmonic compensation algorithm is implemented discretely in the θ domain. The θ domain (0-2π) is discretized into N equal parts with an interval of Therefore, the fundamental period in the angular position domain is N. The value of N depends on the highest harmonic to be compensated and the resolution within one period. For example, to compensate for the highest 128th harmonic, at least 8 points are collected in one harmonic period, and the minimum value of M is 1024;

[0015] A2. Mapping time domain to angular position domain;

[0016] At the mth sampling moment (sampling period is T s ) to obtain the motor rotor angle θ(m) and current error Δi q (m), the mapping from time domain to angular position domain is completed by the following formula.

[0017]

[0018]

[0019] Where round() means rounding to the nearest integer, n is the position domain coordinate, and the value range of n is 0 to N-1;

[0020] When the motor speed is high, the adjacent discrete time domain indexes m-1 and m are mapped to non-adjacent discrete position domain indexes n1 and n2, respectively, where n2-n1≥2;

[0021] In view of the above situation, the angular position domain interpolation update is performed according to the following formula. The following operations on n are all performed on the finite field GF(N), that is, n 1 , n 2 , the value range of n is 0 to N-1. If the result of the operation is not less than N, a modulo N operation is performed on it, and the result after the modulo N operation is taken as the final result; for example: n 1 When the right side of the equation is N, n 1 The actual value will be 0; when the calculation result is -2, n 1 The actual value will be assigned to N-2;

[0022]

[0023]

[0024]

[0025] A3. Synchronous low-pass average filtering;

[0026] exist After updating, the synchronous low-pass average filtering is realized according to the following formula to complete I q Update

[0027]

[0028] A4. Calculation of harmonic components in the angular position domain;

[0029]

[0030]

[0031] Assume that there is no harmonic compensator qin to Δi q The transfer function is Q(z), then

[0032]

[0033]

[0034] Where T s It's to control the beat. and are the amplitude and phase of the transfer function Q(z) at the harmonic frequency point kω respectively;

[0035] A5. Adaptive adjustment of parameters;

[0036] The harmonic components in the speed error are averaged in the angular position domain, and the amplitude of the harmonic compensation is adjusted and updated as follows:

[0037]

[0038]

[0039] Where j is the number of iterations, g k is the adaptive adjustment gain. The update cycle of the above parameters is the time required for the motor to rotate r times. The faster the motor speed, the shorter the update cycle. Therefore, when the motor speed changes, the adaptive adjustment is a time-varying control system in the time domain.

[0040] A6. Calculate the sampling time k th Harmonic compensation amount;

[0041]

[0042] A7. Obtain the injected signal;

[0043] In each current loop control cycle, the injected current is updated.

[0044] Preferably, in the above 5 steps, k Design according to the following method:

[0045] Take q axis k th Taking harmonic compensation as an example, we study qin to a qk (j) and i qin to a qk (j) System response;

[0046] Assume that the q-axis PI controller injects current i without harmonic compensator qinj To q-axis current deviation Δi q The transfer function Q(z);

[0047] And assume:

[0048]

[0049]

[0050]

[0051] Then we have:

[0052]

[0053] After conversion from time domain to position domain, we have:

[0054]

[0055] Further we can get:

[0056]

[0057]

[0058] From the above, it can be seen that the average in the position domain is equivalent to a DFT of one circle in the position domain. The calculation simplification can be obtained:

[0059]

[0060] Before the outer harmonic compensation loop parameters are updated, and have gradually converged to the amplitude of the harmonic component corresponding to the Q(z) output. Before the adaptive iterative update, a qk (j), b qk (j) have converged to i qin The amplitude of the corresponding harmonic component;

[0061] Get the equivalent k th Harmonic compensation.

[0062] Therefore, from k th From the perspective of the transmission of harmonics in the control loop, The adaptive adjustment is equivalent to Figure 4 shown.

[0063] By A k arrive or B k arrive The transfer function is:

[0064]

[0065] The condition for system stability is that the poles of the transfer function are within the unit circle, so 0 < g k <2. In particular, take g k =1, under ideal conditions, convergence can be achieved after one iteration.

[0066] Preferably, in the above step A7, multiple harmonics can be compensated at the same time. After calculating the compensation amount of each harmonic according to the method described in the previous section, add them and inject them into the input of the current loop controller; if k~l (l≥k) harmonic compensation is performed at the same time, in each current loop control cycle, the injected current can be updated according to the following formula:

[0067]

[0068] The beneficial effects of the present invention are as follows: the harmonics of the phase current can be compensated in the dq coordinate system, and the d-axis current feedback value i d and q-axis current feedback value i q Medium harmonics, to suppress the motor U-phase feedback current i in the three-phase current u , motor V phase feedback current i v and the motor W phase feedback current i w The purpose of harmonics. It has the following beneficial effects: compensating the harmonics of phase current in the dq coordinate system; performing harmonic identification and compensation in the motor rotor angle position domain; working effectively at different motor speeds; multiple harmonic components can be compensated simultaneously without coupling with each other; using a synchronous low-pass filter in the position domain to reduce the computational complexity while reducing the impact of noise on harmonic identification; the outer loop parameters are adaptively iterated, and the inner loop feedforward compensation ensures the original robustness and control performance of the current loop. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] Figure 1 The present invention is a schematic diagram of a servo current loop control block diagram of an adaptive controller for suppressing current harmonics with current harmonic suppression.

[0070] Figure 2k is the adaptive controller for suppressing current harmonics in the present invention. th Schematic diagram of the principle structure of the harmonic compensator.

[0071] Figure 3 The present invention is a schematic diagram of an equivalent block diagram of a current harmonic suppressor in an adaptive controller for suppressing current harmonics.

[0072] Figure 4 k is the adaptive controller for suppressing current harmonics in the present invention. th Schematic diagram of the equivalent diagram of the harmonic compensator.

[0073] Figure 5 The invention is a schematic diagram of a block diagram of obtaining an injection signal in an adaptive controller for suppressing current harmonics.

[0074] Figure 6 It is a schematic diagram of the U-phase current power spectrum of the permanent magnet synchronous motor in the prior art when the current harmonics are suppressed and the motor is running at 3000 rpm.

[0075] Figure 7 It is a schematic diagram of dq axis current waveforms when the adaptive controller for suppressing current harmonics of the present invention does not use harmonic compensation in the position domain.

[0076] Figure 8 It is a schematic diagram of the dq axis current spectrum of the adaptive controller for suppressing current harmonics of the present invention when harmonic compensation is not used in the position domain.

[0077] Fig. 9 It is a schematic diagram of the dq axis current waveform of the adaptive controller for suppressing current harmonics of the present invention using harmonic compensation in the current loop in the position domain to compensate for 24th and 48th harmonics.

[0078] Fig.10 It is a schematic diagram of the dq axis current spectrum of the adaptive controller for suppressing current harmonics of the present invention, which uses harmonic compensation in the current loop in the position domain to compensate for 24th and 48th harmonics.

[0079] Fig.11 It is a schematic diagram of phase current waveforms when the adaptive controller for suppressing current harmonics of the present invention does not use harmonic compensation in the position domain.

[0080] Fig.12 It is a schematic diagram of the phase current spectrum when the adaptive controller for suppressing current harmonics of the present invention does not use harmonic compensation in the position domain.

[0081] Fig.13 It is a schematic diagram of the phase current waveform of the adaptive controller for suppressing current harmonics of the present invention using harmonic compensation in the current loop in the position domain to compensate for 24th and 48th harmonics.

[0082] Fig.14It is a schematic diagram of the phase current spectrum of 24th and 48th harmonics in which the adaptive controller for suppressing current harmonics of the present invention uses harmonic compensation in the current loop in the position domain.

[0083] Fig.15 It is a schematic diagram of phase current waveforms when the adaptive controller for suppressing current harmonics of the present invention does not use harmonic compensation in the time domain.

[0084] Fig.16 It is a schematic diagram of the phase current spectrum of the adaptive controller for suppressing current harmonics of the present invention when harmonic compensation is not used in the time domain.

[0085] Fig.17 It is a schematic diagram of the phase current waveform of the adaptive controller for suppressing current harmonics of the present invention, which uses harmonic compensation in the current loop in the time domain to compensate for 24th and 48th harmonics.

[0086] Fig.18 It is a schematic diagram of the phase current spectrum of 24th and 48th harmonics compensated by using harmonic compensation in the current loop of the adaptive controller for suppressing current harmonics of the present invention in the time domain. DETAILED DESCRIPTION

[0087] The present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0088] Figure 1 , Figure 2 , Figure 3 , Figure 4 , Figure 5 In the embodiment shown, an adaptive controller for suppressing current harmonics includes a Park inverter, a Park converter, a controlled object P(Z), a d-axis proportional integral (PI) controller PI 1 , q-axis PI controller PI 2 , Clark converter, d-axis current deviation Δi d With the d-axis injection current i dinj The differential output d-axis PI controller input current i din To d-axis PI controller PI 1 , q-axis current deviation Δi q With the q-axis injection current i qinj The differential output q-axis PI controller input current i qin To q-axis PI controller PI 2 , P.I. 1 The output d-axis control voltage u d ,PI 2 The output q-axis control voltage u q ,d-axis control voltage u d and q-axis control voltage u qAfter the park inverse converter, the α-axis control voltage u is output α and β-axis control voltage u β , input the controlled object P(Z), the controlled object P(Z) feeds back the motor U phase feedback current i in the three-phase current of the motor u , motor V phase feedback current i v and the motor W phase feedback current i w To the Clark converter, after being processed and transformed by the Clark converter, the α-axis feedback current i is output α and β-axis feedback current i β To the park converter, after the park converter processing and transformation, the d-axis current feedback value i is output respectively. d and q-axis current feedback value i q , where the d-axis feedback current i d With d-axis current given value The differential output d-axis current deviation Δi d , q-axis current feedback value i q With q-axis current given value The differential output q-axis current deviation Δi q ; d-axis current deviation Δi d and q-axis current deviation Δi q Input to the time domain → angular position domain mapper, after being processed by the time domain → angular position domain mapper, the d-axis current deviation with position domain coordinate n is output and the q-axis current deviation with position domain coordinate n Then it is input into the synchronous low-pass filter, and after filtering by the synchronous low-pass filter, the output is the d-axis current deviation I with the filtered position domain coordinate n. d (n) and the q-axis current deviation I with the filtered position domain coordinate n q (n) to k th Harmonic compensator, k th The harmonic compensator outputs the d-axis PI controller injection current i dinj and the q-axis PI controller injects current i qinj ; By suppressing the d-axis current feedback value i d and q-axis current feedback value i q Medium harmonics, to suppress the motor U-phase feedback current i in the three-phase current u , motor V phase feedback current i v and the motor W phase feedback current i w The purpose of harmonics. The harmonics of the phase current can be compensated in the dq coordinate system by suppressing the d-axis current feedback value i d and q-axis current feedback value i q Medium harmonics, to suppress the motor U-phase feedback current i in the three-phase current u , motor V phase feedback current iv and the motor W phase feedback current i w The purpose of harmonics.

[0089] In steady-state operation, the d-axis reference With q-axis reference is a constant, so i d with i q The harmonic components in the current error Δi d , Δi q Therefore, by extracting Δi d , Δi q The harmonic components in the circuit are adjusted by the adaptive harmonic compensation module to adjust the amplitude of each injected harmonic. Finally, the injected currents of each harmonic compensation module are added together and injected into the input of the current loop controller to minimize Δi d , Δi q The purpose of each harmonic content is to suppress the corresponding harmonic components in the phase current.

[0090] k th The harmonic compensator includes a first adaptive iterator corresponding to the d-axis and a second adaptive iterator corresponding to the q-axis, forming a d-axis k th Harmonic compensator and q-axis k th Harmonic compensator, where d axis k th Harmonic compensator and q-axis k th The harmonic compensators all use the same corresponding algorithm implementation steps, where the q axis k th The harmonic compensator adopts the following algorithm implementation steps:

[0091] A1. Discretization of angular position domain;

[0092] The harmonic compensation algorithm is implemented discretely in the θ domain. The θ domain (0-2π) is discretized into N equal parts with an interval of Therefore, the fundamental period in the angular position domain is N. The value of N depends on the highest harmonic to be compensated and the resolution within one period. For example, to compensate for the highest 128th harmonic, at least 8 points are collected in one harmonic period, and the minimum value of M is 1024;

[0093] A2. Mapping time domain to angular position domain;

[0094] At the mth sampling moment (sampling period is T s ) to obtain the motor rotor angle θ(m) and current error Δi q (m), the mapping from time domain to angular position domain is completed by the following formula.

[0095]

[0096]

[0097] Where round() means rounding to the nearest integer, n is the position domain coordinate, and the value range of n is 0 to N-1; when the motor speed is high, the adjacent discrete time domain indexes m-1 and m will be mapped to non-adjacent discrete position domain indexes n respectively. 1 With n 2 , where n 2 -n 1≥ 2;

[0098] In view of the above situation, the angular position domain interpolation update is performed according to the following formula. The following operations on n are all performed on the finite field GF(N), that is, n 1 , n 2 , the value range of n is 0 to N-1. If the result of the operation is not less than N, a modulo N operation is performed on it, and the result after the modulo N operation is taken as the final result; for example: n 1 When the right side of the equation is N, n 1 The actual value will be 0; when the calculation result is -2, n 1 The actual value will be assigned to N-2;

[0099]

[0100]

[0101]

[0102] A3. Synchronous low-pass average filtering;

[0103] exist After updating, the synchronous low-pass average filtering is realized according to the following formula to complete I q Update

[0104]

[0105] A4. Calculation of harmonic components in the angular position domain;

[0106]

[0107]

[0108] exist Figure 2 In the example, it is assumed that there is no harmonic compensator. qin to Δi q The transfer function is Q(z), then

[0109]

[0110]

[0111] Where Ts It is to control the beat. and are the amplitude and phase of the transfer function Q(z) at the harmonic frequency point kω respectively;

[0112] A5. Adaptive adjustment of parameters;

[0113] The harmonic components in the speed error are averaged in the angular position domain, and the amplitude of the harmonic compensation is adjusted and updated as follows:

[0114]

[0115]

[0116] Where j is the number of iterations, g k is the adaptive adjustment gain. The update cycle of the above parameters is the time required for the motor to rotate r times. The faster the motor speed, the shorter the update cycle. Therefore, when the motor speed changes, the adaptive adjustment is a time-varying control system in the time domain.

[0117] A6. Calculate the sampling time k th Harmonic compensation amount;

[0118]

[0119] A7. Obtain the injected signal;

[0120] In each current loop control cycle, the injected current is updated.

[0121] Preferably, in the above 5 steps, k Design according to the following method:

[0122] Take q axis k th Taking harmonic compensation as an example, we study qin to a qk (j) and i qin to b qk (j) System response;

[0123] Assume that the q-axis PI controller injects current i without harmonic compensator qinj To q-axis current deviation Δi q The transfer function Q(z);

[0124] And assume that:

[0125]

[0126]

[0127]

[0128] Then we have:

[0129]

[0130] After conversion from time domain to position domain, we have:

[0131]

[0132] Further we can get:

[0133]

[0134]

[0135] From the above, it can be seen that the average in the position domain is equivalent to a DFT of one circle in the position domain. The calculation simplification can be obtained:

[0136]

[0137] Before the outer harmonic compensation loop parameters are updated, and have gradually converged to the amplitude of the harmonic component corresponding to the Q(z) output. Before the adaptive iterative update, a qk (j), b qk (j) have converged to i qin The amplitude of the corresponding harmonic component; get the equivalent k th Harmonic compensation.

[0138] Therefore, from k th From the perspective of the transmission of harmonics in the control loop, The adaptive adjustment is equivalent to Figure 4 As shown;

[0139] By A k arrive or B k arrive The transfer function is:

[0140]

[0141] The condition for system stability is that the poles of the transfer function are within the unit circle, so 0 < g k <2. In particular, take g k =1, under ideal conditions, convergence can be achieved after one iteration.

[0142] Preferably, in the above step A8, multiple harmonics can be compensated at the same time. After calculating the compensation amount of each harmonic according to the method described in the previous section, add them and inject them into the input of the current loop controller; if k~l (l≥k) harmonic compensation is performed at the same time, in each current loop control cycle, the injected current can be updated according to the following formula:

[0143]

[0144] A6. Calculate the sampling time k th Harmonic compensation amount;

[0145]

[0146] Figure 1 The variables and their meanings are shown in the following table:

[0147]

[0148] Figure 2 The new variables and their meanings shown in the table below:

[0149]

[0150] Control effect of adaptive controller: 4 pairs of 750W motors are used in the speed loop, and the speed command is set to 3000RPM. 24th harmonic compensation and 48th harmonic compensation are used at the same time. At this time, the base frequency is 3000 / 60=50Hz. Figure 7 , Figure 8 , Fig. 9 , Fig.10 The figure shows the comparison results of the dq axis current algorithm test. Observe the current waveform in the position domain ( Figure 7 and Fig. 9 ), it can be seen that before harmonic compensation, the d-axis fluctuation amplitude and fluctuation frequency are large, and the q-axis fluctuation frequency is high. After using harmonic compensation, the d-axis fluctuation frequency and fluctuation amplitude are reduced, and the q-axis fluctuation frequency is reduced. From the spectrum, when harmonic compensation is not performed ( Figure 8 ), observing the d-axis current, the amplitude of the 24th harmonic is about 290mA, and the amplitude of the 48th harmonic is about 230mA. After using harmonic compensation, the amplitudes of the 24th and 48th harmonics of the d-axis current are both less than 8mA. In addition, other frequency components are not amplified.

[0151] from Fig.11 , Fig.12 , Fig.13 , Fig.14 The figure shows the comparison results of the position domain phase current algorithm test. The 24th harmonic (6th harmonic of current) of the dq axis is presented as the 20th and 28th harmonics (5th and 7th harmonics of current) of the phase current, and the 48th harmonic (12th harmonic of current) of the dq axis is presented as the 44th and 52nd harmonics (11th and 13th harmonics) of the phase current. Observe the current waveform in the position domain ( Fig.11 and Fig.13 ), it can be seen that before harmonic compensation, the phase current has more burrs, and after using harmonic compensation, the current waveform approaches the ideal sine wave; from the spectrum, when no harmonic compensation is performed ( Fig.12 ), observe the U phase current, the amplitude of the 20th harmonic is about 18mA, the amplitude of the 28th harmonic is about 11mA, the amplitude of the 44th harmonic is about 21mA, and the amplitude of the 52nd harmonic is about 7mA. After using harmonic compensation ( Fig.14 ), the amplitudes of the 20th, 28th, 44th and 52nd harmonics of the U-phase current are all less than 4mA, and other frequency components are not amplified.

[0152] from Fig.15 , Fig.16 , Fig.17 , Fig.18 The figure shows the comparison results of the time domain phase current algorithm test. Fig.16 and Fig.18 ) The effect before and after compensation is consistent with the phenomenon of phase current spectrum in the position domain ( Fig.12 and Fig.14 ), after harmonic compensation, the amplitude of the compensated harmonic is significantly reduced, and other harmonic components are not affected. Compared with the above position domain comparison, from the time domain phase current waveform ( Fig.15 and Fig.17 ) cannot directly observe the difference before and after compensation. This also reflects the superiority of angular position domain processing.

[0153] It can be seen that after using the adaptive controller for suppressing current harmonics in the position domain, the purpose of suppressing the corresponding harmonic components in the phase current is achieved by compensating for the harmonic components of specified orders in the dq axis, and at the same time it will not affect the components of other frequencies.

[0154] The above content and structure describe the basic principle, main features and advantages of the product of the present invention, which should be understood by those skilled in the art. The above examples and descriptions are only for explaining the principle of the present invention. Without departing from the spirit and scope of the present invention, the present invention may have various changes and improvements, which are within the scope of the present invention. The scope of the present invention is defined by the attached claims and their equivalents.

Claims

1. An adaptive controller for suppressing current harmonics, characterized in that: Including Park inverse converter, Park converter, controlled object P(Z), d-axis proportional integral (PI) controller PI1, q-axis PI controller PI2, Clark converter, d-axis current deviation Δi d With the d-axis injection current i dinj The differential output d-axis PI controller input current i din To d-axis PI controller PI1, q-axis current deviation Δi q With the q-axis injection current i qinj The differential output q-axis PI controller input current i qin To the q-axis PI controller PI2, the output d-axis control voltage u of PI1 d ,PI2 output q-axis control voltage u q ,d-axis control voltage u d and q-axis control voltage u q After the park inverse converter, the α-axis control voltage u is output α and β-axis control voltage u β , input the controlled object P(Z), the controlled object P(Z) feeds back the motor U phase feedback current i in the three-phase current of the motor u , motor V phase feedback current i v and motor W phase feedback current i w To the Clark converter, after being processed and transformed by the Clark converter, the α-axis feedback current i is output α and β-axis feedback current i β To the park converter, after the park converter processing and transformation, the d-axis current feedback value i is output respectively. d and q-axis current feedback value i q , where the d-axis feedback current i d With d-axis current given value The differential output d-axis current deviation Δi d , q-axis current feedback value i q With q-axis current given value The differential output q-axis current deviation Δi q ; d-axis current deviation Δi d and q-axis current deviation Δi q Input to the time domain → angular position domain mapper, after being processed by the time domain → angular position domain mapper, the d-axis current deviation with position domain coordinate n is output and the q-axis current deviation with position domain coordinate n Then it is input into the synchronous low-pass filter, and after filtering by the synchronous low-pass filter, the output is the d-axis current deviation I with the filtered position domain coordinate n. d (n) and the q-axis current deviation I with the filtered position domain coordinate n q (n) to k th Harmonic compensator, k th The harmonic compensator outputs the d-axis PI controller injection current i dinj and the q-axis PI controller injects current i qinj ; By suppressing the d-axis current feedback value i d and q-axis current feedback value i q Medium harmonics, to suppress the motor U-phase feedback current i in the three-phase current u , motor V phase feedback current i v and motor W phase feedback current i w The purpose of harmonics.

2. The adaptive controller for suppressing current harmonics according to claim 1, characterized in that: The k th The harmonic compensator includes a first adaptive iterator corresponding to the d-axis and a second adaptive iterator corresponding to the q-axis, forming a d-axis k th Harmonic compensator and q-axis k th Harmonic compensator, where d axis k th Harmonic compensator and q-axis k th The harmonic compensators all use the same corresponding algorithm implementation steps, where the q axis k th The harmonic compensator adopts the following algorithm implementation steps: A1. Discretization of angular position domain; The harmonic compensation algorithm is implemented discretely in the θ domain. The θ domain (0-2π) is discretized into N equal parts with an interval of Therefore, the fundamental period in the angular position domain is N; the value of N depends on the highest harmonic to be compensated and the resolution within one period; for example, to compensate for the highest 128th harmonic, at least 8 points are collected in one harmonic period, and the minimum value of M is 1024; A2. Mapping time domain to angular position domain; At the mth sampling moment (sampling period is T s ) to obtain the motor rotor angle θ(m) and current error Δi q (m), the mapping from time domain to angular position domain is completed by the following formula. Where round() means rounding to the nearest integer, n is the position domain coordinate, and the value range of n is 0 to N-1; When the motor speed is high, the adjacent discrete time domain indexes m-1 and m are mapped to non-adjacent discrete position domain indexes n1 and n2 respectively, where n2-n 1≥ 2; In view of the above situation, the angular position domain interpolation update is performed according to the following formula. The following operations on n are all performed on the finite field GF(N), that is, the value range of n1, n2, and n is 0 to N-1. If the operation result is not less than N, a modulo N operation is performed on it, and the result after the modulo N operation is used as the final result; for example: when the calculation result on the right side of the equation for n1 is N, n1 will actually be assigned a value of 0; when the calculation result is -2, n1 will actually be assigned a value of N-2; A3. Synchronous low-pass average filtering; exist After updating, the synchronous low-pass average filtering is realized according to the following formula to complete I q Update A4. Calculation of harmonic components in the angular position domain; Assume that there is no harmonic compensator qin to Δi q The transfer function is Q(z), then Where T s It's to control the beat. and are the amplitude and phase of the transfer function Q(z) at the harmonic frequency point kω respectively; A5. Adaptive adjustment of parameters; The harmonic components in the speed error are averaged in the angular position domain, and the amplitude of the harmonic compensation is adjusted and updated as follows: Where j is the number of iterations, g k is the adaptive adjustment gain. The update cycle of the above parameters is the time required for the motor to rotate r times. The faster the motor speed, the shorter the update cycle. Therefore, when the motor speed changes, the adaptive adjustment is a time-varying control system in the time domain. A6. Calculate the sampling time k th Harmonic compensation amount; A7. Obtain the injected signal; In each current loop control cycle, the injected current is updated.

3. The adaptive controller for suppressing current harmonics according to claim 1, characterized in that: g in step A5 above k Design according to the following method: Take q axis k th Taking harmonic compensation as an example, we study qin to a qk (j) and i qin to b qk (j) System response; Assume that the q-axis PI controller injects current i without harmonic compensator qinj To q-axis current deviation Δi q The transfer function Q(z); And assume: Then we have: After conversion from time domain to position domain, we have: Further we can get: From the above, it can be seen that the average in the position domain is equivalent to a DFT of one circle in the position domain. The calculation simplification can be obtained: Before the outer harmonic compensation loop parameters are updated, and have gradually converged to the amplitude of the harmonic component corresponding to the Q(z) output. Before the adaptive iterative update, a qk (j), b qk (j) have converged to i qin The amplitude of the corresponding harmonic component; get the equivalent k th Harmonic compensation.

4. The adaptive controller for suppressing current harmonics according to claim 3, characterized in that: The equivalent k th In harmonic compensation, A k , B k The current signal i qin K th The amplitude of the sine and cosine signals of the harmonics; By A k arrive or B k arrive The transfer function is: The condition for system stability is that the poles of the transfer function are within the unit circle, so 0 < g k <2; in particular, take g k =1, under ideal conditions, convergence can be achieved after one iteration.

5. The adaptive controller for suppressing current harmonics according to claim 1, characterized in that: In the above step A8; Multiple harmonics can be compensated at the same time. After calculating the compensation amount of each harmonic according to the method described in the previous section, add them up and inject them into the input of the current loop controller; If k~l (l≥k) harmonic compensation is performed simultaneously, the injected current can be updated according to the following formula in each current loop control cycle:

Images