High-precision model predictive current control method suitable for dual three-phase permanent magnet synchronous motor

Abstract

The invention discloses a high-precision model predictive current control method suitable for a dual three-phase permanent magnet synchronous motor, which comprises the following steps of: deducing a current discrete model of the dual three-phase permanent magnet synchronous motor by considering the change of a rotor position in a control period; constructing a candidate virtual voltage vector set consisting of 24 virtual voltage vectors and zero vectors, wherein the voltage components of the 24 virtual voltage vectors are 0 on a voltage vector xy plane; taking the error squares of the dq-axis current and the xy-axis current of the dual three-phase permanent magnet synchronous motor at the end of each control cycle as evaluation indexes, constructing a cost function, and evaluating the errors of the dq-axis current and the xy-axis current corresponding to different voltage vectors in the candidate virtual voltage vector set at the end of each control cycle by using the cost function. And screening to obtain a voltage vector with the minimum cost function and outputting the voltage vector. The method has smaller dq-axis current tracking error and electromagnetic torque tracking error when the motor operates at a low carrier wave ratio, and improves the steady-state performance of motor operation.

Description

Technical Field

[0001] The present invention belongs to the technical field of motor control, and in particular relates to a high-precision model prediction current control method suitable for a dual three-phase permanent magnet synchronous motor. Background Art

[0002] The increasing requirements for power density and reliability of motor drive systems in aerospace, ship drive, electric vehicles and other application fields have gradually highlighted the limitations of ordinary three-phase motors, and multi-phase motors have attracted a lot of attention and research. Dual three-phase permanent magnet synchronous motors are one of the most widely studied multi-phase motors. They have the advantages of strong fault tolerance, high power density, and small torque pulsation, and are suitable for high-power traction and transmission systems. In order to reduce system losses, it is often necessary to appropriately reduce the switching frequency of the drive system. However, this will cause the motor to operate under low carrier ratio conditions, reducing the dynamic and steady-state performance of the motor.

[0003] In the process of realizing predictive control of motors through digital controllers, most of the existing predictive current control technologies use the forward Euler approximation method to discretize the motor model, such as references [Y.Luo and C.Liu,"Multi-Vector-Based Model Predictive Torque Control for a Six-Phase PMSM Motor With Fixed Switching Frequency"in IEEE Transactions on Energy Conversion,vol.34,no.3,pp.1369-1379,Sep.2019] and references [S.Liu and C.Liu,"Virtual-Vector-Based Robust Predictive Current Control for Dual Three-Phase PMSM"in IEEE Transactions on Industrial Electronics,vol.68,no.3,pp.2048-2058,Mar.2021], and then perform current prediction based on candidate voltage vectors, traverse and screen the optimal voltage vector according to the cost function, and output it through the inverter. However, the predictive current control technology using the forward Euler approximation method assumes that the rotor position remains unchanged within a control cycle. When the motor speed is low or the control frequency is high, this technology can more accurately predict and track the current; however, as the motor speed increases or the control frequency decreases, the rotor position will change significantly within a control cycle, and the error caused by the discretization of the model will increase accordingly. In addition, the control delay will also increase, resulting in problems such as reduced motor current prediction accuracy and dynamic steady-state performance.

[0004] In summary, it can be seen that the current tracking performance and electromagnetic torque tracking performance of the prior art when the dual three-phase permanent magnet synchronous motor operates under low carrier ratio conditions still need to be improved. Summary of the invention

[0005] In view of the above, the present invention provides a high-precision model predictive current control method suitable for a dual three-phase permanent magnet synchronous motor, which can improve the motor current and electromagnetic torque tracking performance under the motor low carrier ratio operating condition.

[0006] A high-precision model prediction current control method suitable for a dual three-phase permanent magnet synchronous motor comprises the following steps:

[0007] (1) Considering the change of rotor position within a control cycle, a discrete model of dq axis current of dual three-phase permanent magnet synchronous motor is established by combining the free component and forced component of dq plane current;

[0008] (2) Based on the solution method of the first-order linear differential equation, the xy-axis current discrete model of the dual three-phase permanent magnet synchronous motor is established;

[0009] (3) obtaining a candidate voltage vector set by screening and synthesizing all output voltage vectors of the motor inverter;

[0010] (4) Determine the d-axis stator current i in the k+2th control period according to the dq-axis current discrete model d k+2 and the q-axis stator current i q k+2 The calculation expression of k is a natural number. According to the xy-axis current discrete model, the x-axis stator current i of the k+2th control period is determined. x k+2 and the y-axis stator current i y k+2 The computational expression of , and then construct the cost function;

[0011] (5) Substitute the voltage vectors in the candidate voltage vector set into the cost function for calculation, take the voltage vector with the smallest corresponding function value as the optimal voltage vector for the k+1th control cycle, and use the optimal voltage vector to generate a set of PWM (pulse width modulation) signals to control the on and off of the power switching devices in the motor inverter.

[0012] Furthermore, the expression of the dq axis current discrete model is as follows:

[0013]

[0014] Where: i d k+1 and i qk+1 are the d-axis stator current and q-axis stator current of the k+1th control cycle, i d k and i q k are the d-axis stator current and q-axis stator current of the k-th control cycle, ω e is the electrical angular velocity of the motor, ψ f is the permanent magnet flux of the motor, L d and L q are the d-axis stator inductance and q-axis stator inductance of the motor, T s is the control period, u d k and u q k They are the d-axis component and q-axis component of the optimal voltage vector in the kth control cycle respectively.

[0015] Furthermore, the expression of the xy-axis current discrete model is as follows:

[0016]

[0017] Where: i x k+1 and i y k+1 are the x-axis stator current and y-axis stator current of the motor in the k+1th control cycle, i x k and i y k are the x-axis stator current and y-axis stator current of the motor in the kth control cycle, R s is the stator resistance, L σ is the stator leakage inductance, T s is the control period, u x k and u y k are the x-axis component and y-axis component of the optimal voltage vector of the kth control cycle respectively.

[0018] Furthermore, in the step (3), the output voltage vector of the motor inverter includes 60 valid vectors and 4 zero vectors, and the 60 valid vectors are divided into 12 large vectors, 12 sub-large vectors, 24 medium vectors, and 12 small vectors according to the amplitude; large vectors and sub-large vectors with the same direction are selected on the αβ sub-plane to synthesize 12 virtual voltage vectors with a component of 0 on the xy sub-plane, and sub-large vectors and small vectors with the same direction are selected on the αβ sub-plane to synthesize another 12 virtual voltage vectors with a component of 0 on the xy sub-plane, and finally these 24 virtual voltage vectors together with the zero vector constitute a candidate voltage vector set.

[0019] Furthermore, in step (4), d k+2 and i q k+2 The calculation expression is as follows:

[0020]

[0021] Where: i d k+1 and i q k+1 are the d-axis stator current and q-axis stator current of the k+1th control cycle, ω e is the electrical angular velocity of the motor, ψ f is the permanent magnet flux of the motor, L d and L q are the d-axis stator inductance and q-axis stator inductance of the motor, T s is the control period, u d k+1 and u q k+1 are the d-axis component and q-axis component of any voltage vector in the candidate voltage vector set, respectively.

[0022] Furthermore, in step (4), x k+2 and i y k+2 The calculation expression is as follows:

[0023]

[0024] Where: i x k+1 and i y k+1 are the x-axis stator current and y-axis stator current of the motor in the k+1th control cycle, R s is the stator resistance, L σ is the stator leakage inductance, T s is the control period, u x k+1 and u y k+1 are the x-axis component and y-axis component of any voltage vector in the candidate voltage vector set, respectively.

[0025] Furthermore, the cost function is expressed as follows:

[0026]

[0027] Among them: g is the cost function, λ is the weight coefficient, i d * and i q* are the reference values ​​of the d-axis stator current and the q-axis stator current, respectively, i x * and i y * are the reference values ​​of the x-axis stator current and the y-axis stator current respectively.

[0028] Furthermore, the reference value i d * 、i x * and i y * are all set to 0, the reference value i q * The error between the motor speed command value and the actual value is obtained after PI (proportional integral) control.

[0029] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and the processor is used to execute the computer program to implement the above-mentioned high-precision model prediction current control method applicable to a dual three-phase permanent magnet synchronous motor.

[0030] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the high-precision model prediction current control method applicable to a dual three-phase permanent magnet synchronous motor.

[0031] Based on the above technical solution, the present invention has the following beneficial technical effects:

[0032] 1. The present invention takes into account the change of the motor rotor position within a control cycle and derives a current prediction model for a dual three-phase permanent magnet synchronous motor; compared with the traditional current prediction model based on the first-order forward Euler approximation method, the present invention has better dq-axis current tracking performance and electromagnetic torque tracking performance.

[0033] 2. The present invention takes into account the suppression of xy plane current of the dual three-phase permanent magnet synchronous motor, has good versatility, and is applicable to any finite set model predictive control method using a current discrete model in the dual three-phase permanent magnet synchronous motor. BRIEF DESCRIPTION OF THE DRAWINGS

[0034] Figure 1 The present invention provides a block diagram of a dual three-phase permanent magnet synchronous motor control system based on the high-precision model prediction current control method of the present invention.

[0035] Figure 2 Schematic diagram of the distribution of the output voltage vector of the two-level six-phase inverter on the αβ and xy sub-planes, the left side corresponds to the αβ sub-plane, and the right side corresponds to the xy sub-plane.

[0036] Figure 3Schematic diagram of the virtual vector synthesis principle. The left side corresponds to the αβ sub-plane, and the right side corresponds to the xy sub-plane.

[0037] Figure 4 In order to use the high-precision current discrete model of the present invention and the traditional current discrete model in the control period T s = 320μs motor simulation waveform diagram, the left side corresponds to the traditional current discrete model, and the right side corresponds to the high-precision current discrete model of the present invention. DETAILED DESCRIPTION

[0038] In order to describe the present invention more specifically, the technical solution of the present invention is described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0039] The principle process of the high-precision model prediction current control method applicable to dual three-phase permanent magnet synchronous motors of the present invention is as follows:

[0040] (1) Solve the high-precision dq plane current discrete model of dual three-phase motor.

[0041] Since the resistance voltage drop of the dual three-phase permanent magnet synchronous motor is usually very small and the control period is generally much smaller than the electrical time constant, the resistance voltage drop is usually ignored; under this assumption, the dq axis current equation of the dual three-phase permanent magnet synchronous motor is established in the two-phase synchronous rotating coordinate system oriented by the rotor flux:

[0042]

[0043] Where: u j and i j (j=d,q) is the stator voltage and stator current of the motor j axis, L d and L q is the d-axis stator inductance and the q-axis stator inductance, ω e is the motor electrical angular velocity, ψ f is the permanent magnetic flux.

[0044] In the kth control cycle, that is, kT s ≤t≤(k+1)T s , the free component discrete solution of the dq plane current under zero control quantity is obtained as:

[0045]

[0046] Where: T s is the control period, i j k (j=d,q) is kT s The j-axis current at time i j k+1 (j=d,q) is (k+1)T sThe j-axis current at .

[0047] and the control quantity u d and u q The relevant dq plane current forced component solution is:

[0048]

[0049] Where: u j k (j=d,q) is kT s The j-axis voltage at .

[0050] Combining the solutions of the free and forced components of the dq plane current, the high-precision current discrete model of the dual three-phase permanent magnet synchronous motor dq plane can be solved as follows:

[0051]

[0052] (2) Solve the high-precision xy-plane current discrete model of the dual three-phase permanent magnet synchronous motor.

[0053] According to the solution method of the first-order linear differential equation, the discrete solution of the xy-axis current can be obtained as:

[0054]

[0055] Where: R s is the stator resistance, L σ is the stator leakage inductance, i j k and u j k (j=x,y) is kT s The j-axis current and voltage at i j k+1 is (k+1)T s The j-axis current at .

[0056] (3) Construct a set of candidate voltage vectors.

[0057] The voltage vector plane of the dual three-phase permanent magnet synchronous motor driven by a two-level six-phase voltage source inverter has 60 effective vectors and 4 zero vectors, among which the voltage vector distribution of the αβ and xy sub-planes is as follows: Figure 2 As shown, each voltage vector adopts the same voltage as (S A S B S C ) and (S D S E S F ) is numbered by combining two decimal numbers equivalent to the binary number of X(X=A,B,C,D,E,F) represents the switch state of the X bridge arm IGBT. When the upper IGBT of the X bridge arm is turned on and the lower IGBT is turned off, S X =1; otherwise, S X =0; for example, the six-phase bridge arm switch state (S A S B S C S D S E S F ) is (100110). The magnitudes of the four groups of vectors from the outside to the inside of the vector plane are respectively the large vector |V L |=0.644U dc , the next largest vector |V ML |=0.471U dc , medium vector |V M |=0.333U dc and small vector |V S |=0.173U dc , U dc is the DC bus voltage.

[0058] The traditional model predictive current control applied to dual three-phase permanent magnet synchronous motors only considers the voltage vector effect on the αβ sub-plane, but does not consider the voltage vector effect on the xy sub-plane, which will result in large current harmonic components of the 5th and 7th orders. In order to reduce current harmonics, the present invention selects large vectors and sub-large vectors with the same direction on the αβ sub-plane to synthesize 12 virtual voltage vectors, and selects sub-large vectors and small vectors with the same direction to synthesize another 12 virtual voltage vectors; the large vector and sub-large vector with the same direction on the αβ sub-plane have opposite vector directions on the xy sub-plane, and the sub-large vector and small vector on the αβ sub-plane have opposite vector directions on the xy sub-plane. The amplitude of the synthesized voltage vector on the xy sub-plane can be controlled to zero by adjusting the action time of the vector. The virtual vector VV is synthesized with the large vector 44 and the sub-large vector 65 adjacent to the α axis. 1 , the second largest vector 65 and the small vector 56 combine to form a virtual vector VV 13 For example, the principle of virtual vector synthesis is as follows Figure 3 As shown, the relationship between the two vectors and the synthetic virtual vector is:

[0059]

[0060] Where: T s is the control period, t 1 and t 2 They are the large vector V L and the next largest vector V ML The action time, V αβ and Vxy is the component of the synthetic virtual vector in the αβ and xy subplanes. Let |V xy |=0, we can solve for t 1 =0.732T s and t 2 =0.268T s The amplitude of the synthesized virtual vector in the αβ subplane is 0.598U dc Similarly, the action time of the second largest vector and the small vector when synthesizing the virtual vector is 0.578T respectively. s and 0.422T s The amplitude of the synthesized virtual vector in the αβ subplane is 0.345U dc .

[0061] (4) Construct a cost function.

[0062] Considering the delay compensation of predictive current control, the tracking errors of the dq axis and xy axis current are considered in the cost function, and the cost function is expressed as:

[0063]

[0064] Where: i j * (j=d,q,x,y) is the j-axis current reference value, i j k+2 is (k+2)T s The j-axis current at the time, λ is the weight coefficient. Since the candidate voltage vector set is composed of virtual voltage vectors with zero components on the xy plane, the theoretical value of the xy-axis current error term is zero; the current discrete model of the present invention is applicable to any finite set model predictive control method using a current discrete model in a dual three-phase permanent magnet synchronous motor, so this embodiment adopts a general cost function form.

[0065] A dual three-phase permanent magnet synchronous motor control system using the high-precision model prediction current control method of the present invention is as follows: Figure 1 As shown, this method reduces the dq-axis current tracking error and electromagnetic torque tracking error of the dual three-phase permanent magnet synchronous motor under low carrier ratio operating conditions.

[0066] The feasibility and effectiveness of the method of the present invention are verified below in combination with specific simulation data and comparison results.

[0067] In order to verify the feasibility and effectiveness of the method of the present invention, the control strategy based on the high-precision current discrete model is compared with the control strategy based on the traditional current discrete model. The motor parameters are shown in Table 1. The sampling period and carrier period of the control system are the same, which are set to 200μs, 320μs, and 400μs respectively. The control strategies based on the high-precision current discrete model of the present invention and the traditional current discrete model use the same controller parameters. The speed loop uses an anti-integral saturation PI controller, and the proportional, integral, and anti-integral saturation parameters are 0.1885, 1.1844, and 5.3052, respectively.

[0068] Table 1

[0069]

[0070] Figure 4 The waveforms of the two discrete models under load torque step conditions are shown, from top to bottom: electromagnetic torque T e With load torque T L , q-axis current actual value i q act With reference value i q * , actual value of d-axis current i d act With reference value i d * The motor is operated at no load first, and then the load torque jumps from 0 to 8Nm after 1 second, and then jumps to 16Nm after another 1 second.

[0071] Depend on Figure 4 It can be seen that the dq axis current tracking error and electromagnetic torque tracking error of the traditional method are large. This is because the current prediction error of the next moment increases as the control cycle of the traditional method increases. The method of the present invention significantly reduces the tracking error of the dq axis current and electromagnetic torque. When the load torque is 16Nm, the quantitative analysis results of the electromagnetic torque and current are shown in Table 2. The dq axis current and electromagnetic torque tracking errors in Table 2 use the square root mean value σ of the difference between the actual value and the reference value. id , σ iq and σ T The quality of phase current is measured by total harmonic distortion (THD).

[0072] Table 2

[0073]

[0074] It can be seen from Table 2 that when T s = 320μs, under low carrier ratio operating conditions, compared with the method based on the traditional current discrete model, the σ of the method of the present invention is id , σ iqand σ T The reductions were 51.96%, 30.69% and 13.68% respectively, and the reduction effect became more obvious as the control period increased. Meanwhile, the total harmonic distortion rate of the phase current of the method of the present invention was reduced.

[0075] The above description of the embodiments is to facilitate the understanding and application of the present invention by those skilled in the art. It is obvious that those skilled in the art can easily make various modifications to the above embodiments and apply the general principles described herein to other embodiments without creative work. Therefore, the present invention is not limited to the above embodiments. Improvements and modifications made by those skilled in the art to the present invention based on the disclosure of the present invention should be within the protection scope of the present invention.

Claims

1. A high-precision model prediction current control method suitable for a dual three-phase permanent magnet synchronous motor, comprising the following steps: (1) Considering the change of rotor position within a control cycle, a discrete model of dq axis current of dual three-phase permanent magnet synchronous motor is established by combining the free component and forced component of dq plane current; (2) Based on the solution method of the first-order linear differential equation, the xy-axis current discrete model of the dual three-phase permanent magnet synchronous motor is established; (3) obtaining a candidate voltage vector set by screening and synthesizing all output voltage vectors of the motor inverter; (4) Determine the d-axis stator current i in the k+2th control period according to the dq-axis current discrete model d k+2 and the q-axis stator current i q k+2 The calculation expression of k is a natural number. According to the xy-axis current discrete model, the x-axis stator current i of the k+2th control period is determined. x k+2 and the y-axis stator current i y k+2 The computational expression of , and then construct the cost function; (5) Substitute the voltage vectors in the candidate voltage vector set into the cost function for calculation, take the voltage vector with the smallest corresponding function value as the optimal voltage vector for the k+1th control cycle, and use the optimal voltage vector to generate a set of PWM signals to control the on and off of the power switching devices in the motor inverter.

2. The high-precision model prediction current control method according to claim 1, characterized in that: The expression of the dq axis current discrete model is as follows: Where: i d k+1 and i q k+1 are the d-axis stator current and q-axis stator current of the k+1th control cycle, i d k and i q k are the d-axis stator current and q-axis stator current of the k-th control cycle, ω e is the electrical angular velocity of the motor, ψ f is the permanent magnet flux of the motor, L d and L q are the d-axis stator inductance and q-axis stator inductance of the motor, T s is the control period, u d k and u q k They are the d-axis component and q-axis component of the optimal voltage vector in the kth control cycle respectively.

3. The high-precision model prediction current control method according to claim 1, characterized in that: The expression of the xy-axis current discrete model is as follows: Where: i x k+1 and i y k+1 are the x-axis stator current and y-axis stator current of the motor in the k+1th control cycle, i x k and i y k are the x-axis stator current and y-axis stator current of the motor in the kth control cycle, R s is the stator resistance, L σ is the stator leakage inductance, T s is the control period, u x k and u y k are the x-axis component and y-axis component of the optimal voltage vector of the kth control cycle respectively.

4. The high-precision model prediction current control method according to claim 1, characterized in that: In the step (3), the output voltage vector of the motor inverter includes 60 valid vectors and 4 zero vectors, and the 60 valid vectors are divided into 12 large vectors, 12 sub-large vectors, 24 medium vectors, and 12 small vectors according to the amplitude; large vectors and sub-large vectors with the same direction are selected on the αβ sub-plane to synthesize 12 virtual voltage vectors with a component of 0 on the xy sub-plane, and sub-large vectors and small vectors with the same direction are selected on the αβ sub-plane to synthesize another 12 virtual voltage vectors with a component of 0 on the xy sub-plane, and finally these 24 virtual voltage vectors together with the zero vector constitute a candidate voltage vector set.

5. The high-precision model prediction current control method according to claim 1, characterized in that: In step (4), i d k+2 and i q k+2 The calculation expression is as follows: Where: i d k+1 and i q k+1 are the d-axis stator current and q-axis stator current of the k+1th control cycle, ω e is the electrical angular velocity of the motor, ψ f is the permanent magnet flux of the motor, L d and L q are the d-axis stator inductance and q-axis stator inductance of the motor, T s is the control period, u d k+1 and u q k+1 are the d-axis component and q-axis component of any voltage vector in the candidate voltage vector set, respectively.

6. The high-precision model prediction current control method according to claim 1, characterized in that: In step (4), i x k+2 and i y k+2 The calculation expression is as follows: Where: i x k+1 and i y k+1 are the x-axis stator current and y-axis stator current of the motor in the k+1th control cycle, R s is the stator resistance, L σ is the stator leakage inductance, T s is the control period, u x k+1 and u y k+1 are the x-axis component and y-axis component of any voltage vector in the candidate voltage vector set, respectively.

7. The high-precision model prediction current control method according to claim 1, characterized in that: The expression of the cost function is as follows: Among them: g is the cost function, λ is the weight coefficient, i d * and i q * are the reference values ​​of the d-axis stator current and the q-axis stator current, respectively, i x * and i y * are the reference values ​​of the x-axis stator current and the y-axis stator current respectively.

8. The high-precision model prediction current control method according to claim 7, characterized in that: The reference value i d * 、i x * and i y * are all set to 0, the reference value i q * The error between the motor speed command value and the actual value is obtained after PI control.

9. A computer device comprising a memory and a processor, wherein a computer program is stored in the memory, wherein: The processor is used to execute the computer program to implement the high-precision model predictive current control method according to any one of claims 1 to 8.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the high-precision model predictive current control method according to any one of claims 1 to 8 is implemented.

Images