Integrated modulation model predictive control method for dual three-phase permanent magnet motor

Abstract

This invention discloses an integrated modulation model predictive control method for a dual three-phase permanent magnet synchronous motor, comprising: establishing a current prediction model using the duty cycle of the six-phase switching signals of a six-phase inverter as the control quantity; designing a full-speed domain current reference value calculation module combined with a field weakening regulator to generate dq-axis current reference values; using a current prediction error compensator to generate dq-axis current compensation values; establishing a cost function considering the current tracking errors of the dq-axis and xy-axis, and solving for the theoretical optimal values ​​of the duty cycles of phases A, B, D, and E; and establishing a lookup table to solve for the actual optimal six-phase duty cycles from the theoretical optimal values ​​of the duty cycles of phases A, B, D, and E. This invention integrates the control algorithm and modulation algorithm, designing a field weakening regulator that uses the six-phase duty cycle as input, which can improve the dynamic performance and operating efficiency of the dual three-phase permanent magnet synchronous motor.

Description

Technical Field

[0001] This invention belongs to the field of motor control technology, specifically relating to an integrated modulation model predictive control method for a dual three-phase permanent magnet motor. Background Technology

[0002] Dual three-phase permanent magnet synchronous motors offer advantages such as reduced torque ripple, increased power density, and excellent fault tolerance, and are widely used in electric vehicles, electric aircraft, and ship propulsion. These transportation industries require drive motors to operate reliably over a wide speed range; therefore, under a given DC bus voltage constraint, field weakening control and overmodulation techniques are needed to extend the power and speed operating range of the drive motor.

[0003] In the high-speed operating region of the motor, due to the limitation of the inverter's DC bus voltage, the voltage amplitude that the stator winding can generate has an upper limit. When the back electromotive force approaches or exceeds this upper limit as the speed increases, traditional control strategies cannot provide sufficient voltage to ensure accurate current tracking, resulting in a decrease in torque output capability and preventing further speed increases. Furthermore, the commonly used SVPWM (Space Vector Pulse Width Modulation) technology is mainly designed for the linear modulation region and cannot fully utilize the inverter's voltage output capability. To address these issues, researchers have adopted two types of solutions: one is to use field weakening algorithms to reduce the electric drive system's demand on the DC bus voltage, thereby expanding the motor drive system's speed range and improving torque output capability; the other is to utilize OVM (Overmodulation) technology to fully utilize the inverter's voltage output capability.

[0004] Currently, field weakening control methods can be broadly categorized into feedforward and feedback methods. Feedforward methods primarily determine the d- and q-axis reference currents through model calculations and table lookups, exhibiting excellent dynamic performance. However, the accuracy of the field weakening current obtained by this method heavily depends on the flux linkage and inductance parameters, which may vary with motor operating conditions. Feedback methods can use feedback quantities such as terminal voltage, voltage difference before and after the pulse width modulation module, and the error between the PWM (pulse width modulation) period and the duration of the synthesized voltage vector command to track voltage limits, providing better robustness against motor parameter mismatch, but with poorer dynamic performance. Existing field weakening control schemes often incorporate field-oriented controllers (FOCs), employing multiple PI (proportional-integral) controllers in a cascaded control loop to calculate the d- and q-axis currents and voltages. However, the parameters of the PI controllers lack specific physical interpretation and typically require complex tuning schemes. This not only increases the time cost in practical applications but also adds to the burden of optimizing overall control performance for different objectives. Furthermore, handling the saturation problem of coupled systems is also quite complex.

[0005] Overmodulation technology allows the inverter to output voltage outside the linear modulation region. Overmodulation is generally divided into single-mode overmodulation and multi-mode overmodulation. Compared with single-mode overmodulation algorithms, multi-mode overmodulation can improve the voltage output quality of the inverter. However, as long as the output voltage exceeds the linear modulation region, there will be a deviation between the actual output voltage of the inverter and the reference voltage. This will cause the actual control result of the motor to deviate from the calculated result of the control strategy, thus reducing the dynamic response speed of the motor system. Summary of the Invention

[0006] In view of the above, this invention provides an integrated modulation model predictive control method for a dual three-phase permanent magnet motor. The optimal six-phase duty cycles (SPDRs) are directly obtained through optimization calculation of the constructed cost function, eliminating the need for the modulation module's execution step. Furthermore, this invention also designs a field weakening regulator that uses the duty cycle as input, further expanding the motor's speed and torque output range, and the implementation process is simple.

[0007] A model predictive control method for an integrated modulated dual three-phase permanent magnet motor includes the following steps: (1) Construct a current prediction model with the duty cycle of the six-phase switch signal of the motor inverter as the control quantity to predict the stator current of the dq axis and xy axis at the next moment; (2) Design a full-speed domain current reference value calculation module that combines a field weakening regulator, with the duty cycle of the six-phase switch signal and the electromagnetic torque reference value as input, and output the dq axis current reference value; (3) Use the current prediction error compensator to generate the dq axis current compensation value, which is used to compensate for the dq axis stator current prediction value and dq axis current reference value at the next moment; (4) Establish a cost function that considers the current tracking error of the dq axis and xy axis, and take minimizing the cost function as the optimization objective. Solve to obtain the theoretical optimal value of the duty cycle of the four-phase switch signals A, B, D and E under this objective. (5) Based on the vector synthesis relationship and the theoretical optimal value of the duty cycle of the four-phase switch signals A, B, D, and E, the actual optimal value of the duty cycle of the six-phase switch signals is solved, and then the PWM signal is generated by the double zero-sequence voltage injection pulse width modulation method to drive the inverter of the dual three-phase permanent magnet synchronous motor.

[0008] Furthermore, the expression for the current prediction model in step (1) is as follows:

[0009] in: T s For discrete control periods, , , and Corresponding to kTs The actual values ​​of the stator current along the d-axis, q-axis, x-axis, and y-axis at time t. , , and Corresponding to ( k +1) T s Predicted stator current values ​​along the d-axis, q-axis, x-axis, and y-axis at time points. and They are respectively kT s The rotor electrical angular velocity and rotor electrical angle of the motor at constant time, , , , , and Corresponding to kT s The duty cycle of the A~F phase switch signals of the motor inverter at any given time. This is the stator resistance of the motor; and These are the d-axis inductance and q-axis inductance of the motor, respectively. For the stator leakage inductance of the motor, For permanent magnet flux linkage in electric motors, k It is a natural number.

[0010] Furthermore, the expression for the full-speed domain current reference value calculation module in step (2) is as follows:

[0011] in: and These are the current reference values ​​for the d-axis and q-axis, respectively. Given the d-axis current value and Set to 0, The field weakening d-axis current generated by the field weakening regulator. This is the reference value for electromagnetic torque. This represents the number of pole pairs of the motor. and These are the d-axis inductance and q-axis inductance of the motor, respectively. This refers to the permanent magnet flux linkage of the motor.

[0012] Furthermore, the field weakening regulator uses the maximum value of the six-phase switch signal duty cycle minus 1 as the input to the PI controller, and then multiplies the output of the PI controller by the rated current after limiting, thus obtaining the field weakening d-axis current. .

[0013] Furthermore, the expression for the current prediction error compensator in step (3) is as follows:

[0014] in: and These are the current compensation values ​​for the d-axis and q-axis, respectively. and They are respectively kT s The actual values ​​of the stator current along the d-axis and q-axis at time t. and They are ( k +1) T s The predicted stator current values ​​on the d-axis and q-axis at time t, where D() represents the delay function and LPF() represents the low-pass filter function. k It is a natural number.

[0015] Furthermore, the expression for minimizing the cost function in step (4) is as follows:

[0016] in: J Let cost function be and These are the current compensation values ​​for the d-axis and q-axis, respectively. and They are ( k +2) T s Predicted stator current values ​​along the d-axis and q-axis at time t. and They are ( k +2) T s Predicted stator current values ​​along the x and y axes at time points. and These are the current reference values ​​for the d-axis and q-axis, respectively. and These are the current reference values ​​for the x-axis and y-axis, respectively. λ These are the weighting coefficients. , , , They are ( k +1) T s The theoretical optimal duty cycle values ​​for the four-phase switch signals at times A, B, D, and E.

[0017] Furthermore, in step (5), the actual optimal value of the duty cycle of the six-phase switch signal is solved using the following relationship: when d A1 ≥0 and d B1 ≥0 andd E1 <0 and d D1 > d E1 In this case, d A = d A1 , d B = d B1 , d C =0, d D = d D1 - d E1 , d E =0, d F =- d E1 ; when d A1 ≥0 and d B1 ≥0 and d D1 ≥0 and d E1 In the case of ≥0, d A = d A1 , d B = d B1 , d C =0, d D = d D1 , d E = d E1 , d F =0; when d A1 <0 and d A1 ≤ d B1 and d D1 ≥0 and d E1 In the case of ≥0, d A =0, dB = d B1 - d A1 , d C =- d A1 , d D = d D1 , d E = d E1 , d F =0; when d A1 <0 and d A1 ≤ d B1 and d D1 <0 and d D1 ≤ d E1 In this case, d A =0, d B = d B1 - d A1 , d C =- d A1 , d D =0, d E = d E1 - d D1 , d F =- d D1 ; when d B1 <0 and d A1 > d B1 and d D1 <0 and d D1 ≤ d E1 In this case, d A = dA1 - d B1 , d B =0, d C =- d B1 , d D =0, d E = d E1 - d D1 , d F =- d D1 ; when d B1 <0 and d A1 > d B1 and d E1 <0 and d D1 > d E1 In this case, d A = d A1 - d B1 , d B =0, d C =- d B1 , d D = d D1 - d E1 , d E =0, d F =- d E1 ; in: d A1 , d B1 , d D1 , d E1 These are the theoretical optimal values ​​for the duty cycles of the four-phase switch signals A, B, D, and E, respectively. d A , dB , d C , d D , d E , d F These are the actual optimal values ​​for the duty cycle of the six-phase switch signals A, B, C, D, E, and F, respectively.

[0018] A computer device includes a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the above-described integrated modulation model predictive control method for a dual three-phase permanent magnet motor.

[0019] A computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-described integrated modulation model predictive control method for a dual three-phase permanent magnet motor.

[0020] Based on the above technical solution, the present invention has the following beneficial technical effects: 1. Unlike the traditional method of first calculating the reference voltage through the control algorithm and then calculating the six-phase duty cycle according to the modulation algorithm in a cascaded structure, this invention integrates the function of the modulation algorithm into the control algorithm. The six-phase duty cycle can be obtained directly through the control algorithm, thus eliminating the need for an additional modulation algorithm.

[0021] 2. This invention integrates the control algorithm and modulation algorithm into a single design, ensuring that the actual output voltage of the inverter is consistent with the reference voltage required for control, thereby enabling the motor to obtain a better dynamic current response.

[0022] 3. The field weakening controller designed in this invention uses the six-phase duty cycle as the input value and a constant 1 as the limit value of the duty cycle. It is simple and easy to implement, realizes the unification of the full-speed domain control strategy, and reduces the stator current amplitude of the motor under field weakening conditions, thereby improving the operating efficiency of the motor system. Attached Figure Description

[0023] Figure 1 This is the overall control block diagram of the predictive control method for the dual three-phase permanent magnet motor model of the present invention.

[0024] Figure 2 This is a schematic diagram of the voltage vector distribution of a two-level six-phase voltage source inverter. In the diagram, (a) corresponds to the αβ subplane and (b) corresponds to the xy subplane.

[0025] Figure 3 This is a schematic diagram of the basic voltage vector and reference voltage vector selected in this invention.

[0026] Figure 4This is a schematic diagram illustrating the calculation principle of the six-phase duty cycle in this invention. In the figure, (a) shows a schematic diagram of sector division, and (b) corresponds to... d A1 ≥0 and d B1 For calculations ≥0, (c) corresponds to... d A1 <0 and d A1 < d B1 The calculation results, (d) correspond to d B1 <0 and d A1 > d B1 The calculation results.

[0027] Figure 5 This is a schematic diagram illustrating the principle of calculating the full-speed-domain current reference value in conjunction with the field weakening regulator of this invention.

[0028] Figure 6 This is the current prediction error compensator based on a low-pass filter in this invention.

[0029] Figure 7 The diagram shows the experimental results of transient performance comparison of a dual three-phase PMSM drive system when the torque jumps from 2 Nm to 15 Nm at 500 r / min. In the diagram, (a) corresponds to the traditional deadbeat current prediction control strategy, (b) corresponds to the deadbeat current prediction control strategy using the overmodulation method, and (c) corresponds to the control strategy of the present invention.

[0030] Figure 8 The diagram shows the experimental results of transient performance comparison of a dual three-phase PMSM drive system when the torque jumps from 2 Nm to 15 Nm at a speed of 1400 r / min. In the diagram, (a) corresponds to the traditional deadbeat current prediction control strategy, (b) corresponds to the deadbeat current prediction control strategy using the overmodulation method, and (c) corresponds to the control strategy of the present invention.

[0031] Figure 9 The diagram shows the experimental results of transient performance comparison of a dual three-phase PMSM drive system when the torque jumps from 2 Nm to 10 Nm at 1800 r / min. In the diagram, (a) corresponds to the traditional deadbeat current prediction control strategy, (b) corresponds to the deadbeat current prediction control strategy using the overmodulation method, and (c) corresponds to the control strategy of the present invention.

[0032] Figure 10The diagram shows the experimental results of steady-state performance comparison of a dual three-phase PMSM drive system at a speed of 1400 r / min and a torque of 15 Nm. In the diagram, (a) corresponds to the traditional deadbeat current prediction control strategy, (b) corresponds to the deadbeat current prediction control strategy using the overmodulation method, and (c) corresponds to the control strategy of the present invention.

[0033] Figure 11 The diagram shows the experimental results of steady-state performance comparison of a dual three-phase PMSM drive system at 1800 r / min speed and 10 Nm torque. In the diagram, (a) corresponds to the traditional deadbeat current prediction control strategy, (b) corresponds to the deadbeat current prediction control strategy using the overmodulation method, and (c) corresponds to the control strategy of the present invention. Detailed Implementation

[0034] To describe the present invention in more detail, the technical solution of the present invention will be described in detail below with reference to the accompanying drawings and specific embodiments.

[0035] like Figure 1 As shown in the figure, this embodiment provides an integrated modulation dual three-phase permanent magnet motor model predictive control method, and the specific implementation process is as follows: (1) Establish a current prediction model with the duty cycle of the six-phase switching signals of the six-phase inverter as the control quantity.

[0036] Assume the switching combination of the six-phase inverter is as follows: S =[ S A , S B , S C , S D , S E , S F ],when x When the upper bridge arm MOSFET is turned on and the lower bridge arm MOSFET is turned off S x =1; conversely, S x =0, where x =A, B, C, D, E, F. The voltage vector distribution of a two-level six-phase voltage source inverter is as follows: Figure 2 As shown, 64 voltage vectors can be generated in each subspace, including large vectors. V L (0.644 U dc ), medium and large vectors V ML (0.471 U dc), medium vector V M (0.333 U dc ), small vector V S (0.173 U dc and zero vector V 00 , V 07 , V 70 , V 77 Each voltage vector number is represented by two decimal numbers, each corresponding to a switch combination represented by two three-bit binary numbers. S ,like V 64 Corresponding to [1, 1, 0, 1, 0, 0].

[0037] In the αβ subplane, for V 01 , V 10 , V 02 , V 20 , V 04 , V 40 Of the six voltage vectors and the six switching states of the corresponding switch combinations, only one switching state is 1, while the other five switching states are 0. This exhibits certain special characteristics, for example... V 01 switch combination S =[0, 0, 0, 0, 0, 1] only S F Let it be 1. For ease of narration, let's say... V 40 , V 04 , V 20 , V 02 , V 10 , V 01 Rename them respectively U 1~ U 6. The zero vector V 00 Renamed U 0, such as Figure 3 As shown, U 1~ U The expression for 6 in the αβ coordinate system is:

[0038] in: U dc Indicates the DC bus voltage. S Ai , S Bi , S Ci , S Di , S Ei and S Fi They are respectively U i The corresponding A, B, C, D, E, and F phase switch signals, i =1,2,...,6.

[0039] In the voltage vector synthesis process of traditional four-vector SVPWM for dual three-phase permanent magnet synchronous motors, the reference voltage vector is synthesized from four adjacent effective vectors and the zero vector. Figure 3 In u s For example, if only select U 1~ U 6 serves as the basic effective voltage vector. u s can be U 1. U 2. U 3. U 4. U 0 synthesis, as shown in the following formula:

[0040] in: d j ( j =0,1,...,4) respectively represent U j The duty cycle, and 0 ≤ d j ≤1, 0≤ d 1+ d 2+ d 3+ d 4+ d 0≤1; Here, only 0≤1 is used. V 00 As the zero vector.

[0041] Combining the two formulas above, we can obtain:

[0042]

[0043]

[0044] Six-phase duty cycle d A ~ d F The definition is shown in the above formula, which represents the duty cycle of the six-phase switching signal of the inverter corresponding to the reference voltage vector, where U The six-phase switch signals corresponding to 0 are all 0, which can be ignored here. Since 0 ≤ d 1+ d 2+ d 3+ d 4+ d 0≤1 and S Ai , S Bi , S Ci , S , S Ei and S Fi The value of can only be 0 or 1, therefore 0 ≤ d x ≤1, x =A, B, C, D, E, F.

[0045] From the above derivation process, it can be seen that the voltage vector in the voltage plane u s The expression in the α-β coordinate system can be uniformly expressed as:

[0046] in: U A = U dc e j0 / 3, U B = U dc e j2π / 3 / 3, U C = U dc e j4π / 3 / 3, U D = U dc e jπ / 6 / 3, U E = U dc e j5π / 6 / 3, UF = U dc e j3π / 2 / 3, U A ~ U F like Figure 4 As shown in (a), this formula establishes SPDRs ( d A , d B , d C , d D , d E , d F ) and inverter output voltage u s The direct mathematical relationship between them.

[0047] Based on the switching signals corresponding to the six effective voltage vectors and the zero vector selected above, it can be seen that when synthesizing the required reference voltage vector using four adjacent non-zero vectors and the zero vector, there is always one phase in each of the two sets of three-phase windings with a switching signal of 0, i.e. d A , d B , d C At least one of them is 0. d D , d E , d F At least one of them is 0, which creates conditions for simplifying subsequent calculations.

[0048] The dq and xy axis stator currents are in the first... k The predicted value for +1 control cycle is:

[0049] in: , , , , , , and They are respectively kT s Stator voltage and stator current at times d, q, x, and y axes. T s For discrete control cycles.

[0050] Combining the above two equations, the current prediction model expression for a dual three-phase permanent magnet motor with SPDRs as the control variables can be obtained as follows:

[0051] (2) Design a full-speed domain current reference value calculation module that combines a field weakening regulator.

[0052] When the motor operates in the high-speed range, the optimal six-phase duty cycle obtained by the control algorithm will exceed its limit of 1, which will cause the motor current and other state variables to fail to track their reference values. This invention designs a simple and practical field-weakening controller to adjust the d-axis reference current, enabling the optimal six-phase duty cycle to return to its limit value. Figure 5 As shown, the field weakening controller uses 1 minus the optimal six-phase duty cycle with the largest value as the input to the PI regulator. Then, the output of the PI regulator is multiplied by the rated current to obtain the compensation value of the d-axis reference current. i dw This invention achieves field weakening control. The calculation process of the input to the field weakening controller and the implementation of the PI regulator are very simple, making it a simple and efficient field weakening method.

[0053] The formula for calculating the dq-axis current reference value is as follows:

[0054] in: and These are the reference values ​​for the d-axis and q-axis currents. The given value for the d-axis current is given in this example. Set to 0; The field weakening d-axis current generated by the field weakening regulator; This is a reference value for electromagnetic torque. This represents the number of pole pairs of the motor.

[0055] (3) Design a current prediction error compensator.

[0056] To suppress the impact of current errors caused by factors such as motor parameter mismatch and inverter nonlinearity on control performance, this invention employs a current error compensator based on an LPF (low-pass filter), such as... Figure 6 As shown, the compensator will... k The actual dq-axis current within the first control cycle and the current after the delay module. k The predicted dq-axis current over +1 control cycle is used as the input to a low-pass filter, and then the output of the low-pass filter is... and Compensation up to the k +1 control cycle current reference value ( , ) and current prediction ( , )middle.

[0057] (4) Establish a cost function that considers the tracking errors of the dq-axis current and xy-axis current, and solve for the theoretical optimal values ​​of the duty cycle of the four phases A, B, D, and E. d A1 , d B1 , d D1 , d E1 .

[0058] The current prediction model established in this invention contains 6 control variables. d A , d B , d C , d D , d E , d F If this model is adopted, the optimization process of MPC (Model Predictive Control) requires solving for six optimal control inputs, which increases the computational complexity of the predictive control algorithm. As the above analysis shows, within each control cycle... d A , d B , d C At least one of them is 0. d D , d E , d F At least one of them is 0. Therefore, we first assume d C and d F The value is 0, and then the theoretical values ​​of the optimal duty cycles of phases A, B, D, and E under this assumption are calculated based on the preset target. d A1 , d B1 , d D1 , d E1 Considering delay compensation, a cost function is constructed, and then a design is developed based on ( k +1) T s The optimization problem for the six-phase duty cycle as the control input is:

[0059] in: J Represents the cost function. , , , These are the reference values ​​for the d, q, x, and y-axis currents, respectively. and These represent the compensation amounts for d-axis and q-axis current disturbances caused by factors such as motor parameter mismatch. λ These are the weighting coefficients; in this example λ =1, and Set all to 0.

[0060] (5) Establish the theoretical optimal value of the duty cycle of the four phases A, B, D, and E. d A1 , d B1 , d D1 , d E1 Solve for the actual optimal six-phase duty cycle. d A , d B , d C , d D , d E , d F The lookup table for ).

[0061] Since the phase with a zero duty cycle is an arbitrary assumption, the theoretical optimal value obtained by solving is... d A1 , d B1 , d D1 , d E1 It's possible for the duty cycle to be less than 0, while the actual optimal duty cycle is greater than or equal to 0. Further analysis of the actual optimal SPDRs is needed. d A , d B , d C , d D , d E , d F ) and the theoretical value of duty cycle obtained by the optimization algorithm ( d A1 , d B1 ,d D1 , d E1 The relationship between ).

[0062] First, based on the solution obtained... d A1 , d B1 , d D1 , d E1 The positive and negative signs and their relative magnitudes divide the voltage plane into, for example, Figure 4 The six sectors (I-VI) are shown in (a). For the reference voltage vector... u s It can be considered as U A , U B , U C To synthesize u s1 , U D , U E , U F To synthesize u s2 ,Then u s1 and u s2 synthesis u s For ease of calculation, the calculation process is divided into two groups. The first group is based on... d A1 , d B1 calculate d A , d B , d C The second group according to d D1 , d E1 calculate d D , d E , d F The two sets of analysis and calculation principles are similar. Next, we will only take the first set as an example to analyze the calculation of the optimal six-phase duty cycle in three cases: Scenario 1: When d A1 ≥0、 dB1 When ≥0, then u s1 Within sector I or II, such as Figure 4 The green shaded area shown in (b) d C The assumption that = 0 holds true. At this point, d A = d A1 , d B = d B1 , d C =0.

[0063] Scenario 2: When d A1 <0 and d A1 < d B1 Since the actual duty cycle cannot be less than 0, it can be determined that... d C The assumption that =0 is false. Figure 4 From the vector composition relationship in (c), we can see that... u s1 Located within sector III or IV, and should be controlled by U B and U C Synthesis, in fact should be d A The value is 0. Next, we need to use the result obtained from the first step of optimization... d A1 and d B1 ,calculate d B , d C The optimal value. (From) Figure 4 The geometric relationship in (c) can be obtained as follows: d A1 <0 and d B1 Taking the case where >0 as an example, it can be seen that |QM|=|QR|=|ON|=|OP|=| d A1 U A Therefore, | d B U B |=|OM|=|OQ|+|QM|=| d B1 UB |+| d A1 U A |,| d C U C |=|ON|=|OP|=| d A1 U A |. Because of | U A |=| U B |=| U C |and d A1 <0, based on the above analysis, we get d A =0, d B = d B1 - d A1 , d C =- d A1 .

[0064] Scenario 3: When d B1 <0 and d A1 > d B1 At that time, it can be judged d C The assumption that =0 is false. Figure 4 From the vector composition relationship of (d), we can see that u s1 Located within sector V or VI, and has d B =0. From Figure 4 The geometric relationship of (d) can be obtained by using d B1 <0 and d A1 Taking the case where >0 as an example, it can be seen that |PM|=|PR|=|ON|=|OQ|=| d B1 U B Therefore, | d A U A |=|OM|=|OP|+|PM|=| d A1U A |+| d B1 U B |,| d C U C |=|ON|=|OQ|=| d B1 U B |. Because of | U A |=| U B |=| U C |and d B1 <0, based on the above analysis, we get d A = d A1 - d B1 , d B =0, d C =- d B1 .

[0065] Similarly, it can be based on d D1 , d E1 Solve d D , d E , d F In conclusion, under the assumption... d C and d F Optimal calculation obtained when the value is 0 d A1 , d B1 , d D1 , d E1 Then, synthesize the vector. u s The actual optimal six-phase duty cycle for different sectors is shown in Table 1. Using the relationships in Table 1, the results obtained from the first step of optimization can be calculated easily and quickly. d A1 , d B1 , d D1, d E1 The optimal six-phase duty cycle was obtained by looking up the table.

[0066] Table 1

[0067] Finally, a dual zero-sequence voltage injection PWM modulation method is adopted to generate corresponding pulse width modulation PWM signals based on the optimal six-phase duty cycle, so as to drive the dual three-phase permanent magnet synchronous motors.

[0068] To verify the feasibility and effectiveness of the method of this invention, we conducted experiments in a dual three-phase PMSM (Permanent Magnet Synchronous Motor) drive system. The parameters of the dual three-phase motor system used in the experiment are shown in Table 2. The control algorithm of the drive system was executed on a control board based on a DSP (TMS320F28379D) and an FPGA (5CEFA7F27I7N). The sampling frequency and control frequency in the experiment were both 10kHz. The experiment will compare the dynamic and steady-state performance, motor system operating efficiency, etc., of the predictive current control method using a traditional field weakening regulator and the predictive current control method (SPDR-PCC) of this invention. For the predictive current control method using a traditional field weakening regulator, experiments were conducted to verify the scheme using only a dual zero-sequence voltage injection PWM strategy (PWM-PCC) and the scheme using dual zero-sequence voltage injection PWM + overmodulation strategy (OVM-PCC).

[0069] Table 2

[0070] Figures 7-9 The dynamic response waveforms of the d-axis and q-axis currents and phase currents during sudden changes in reference torque at 500 r / min, 1400 r / min, and 1800 r / min are presented for the predictive current control method using a traditional field weakening regulator and the MPC method of this invention. All MPC methods employ... Figure 6 The compensator compensates for current prediction errors. In the experiment, when the motor was running at 500 r / min and 1400 r / min, the reference torque abruptly changed from 2 Nm to 15 Nm; when the motor was running at 1800 r / min, the reference torque abruptly changed from 2 Nm to 10 Nm. Figures 7-9The experimental results show that: ① When the rotational speed is 500 r / min, the q-axis current dynamic response time of the three methods, PWM-PCC, OVM-PCC, and SPDR-PCC, is 0.8 ms. ② When the rotational speed is 1400 r / min, the q-axis current dynamic response times of PWM-PCC, OVM-PCC, and SPDR-PCC are 5.78 ms, 4.81 ms, and 1.90 ms, respectively. ③ When the rotational speed is 1800 r / min, the q-axis current dynamic response times of PWM-PCC, OVM-PCC, and SPDR-PCC are 30.19 ms, 24.68 ms, and 4.15 ms, respectively.

[0071] The above phenomena are analyzed and summarized as follows: ① At a speed of 500 r / min, the optimal voltage vector amplitude found by the MPC algorithm is relatively small and does not exceed the linear modulation region. Therefore, the three MPC methods can achieve the same current dynamic adjustment speed. ② When the speed reaches 1400 r / min or even higher, the optimal voltage vector calculated by the MPC algorithm during the transient process will exceed the voltage limit. Compared with PWM-PCC, OVM-PCC and SPDR-PCC can improve the inverter's voltage output capability, thus achieving better current dynamic response. In addition, unlike OVM-PCC, SPDR-PCC can ensure that the actual output voltage of the inverter is consistent with the reference voltage calculated by the control algorithm. Therefore, SPDR-PCC has the shortest dynamic response time. ③ Under the same field weakening conditions, the amplitude of the d-axis current reference value required by PWM-PCC, OVM-PCC, and SPDR-PCC decreases sequentially. When the d-axis current reference value is smaller, the d-axis current can track the reference value faster.

[0072] Figure 10 and Figure 11 The steady-state experimental waveforms of PWM-PCC, OVM-PCC, and SPDR-PCC under two operating conditions—1400 r / min and 15 Nm torque, and 1800 r / min and 10 Nm torque—are presented, along with the mean values ​​of the d-axis and q-axis current vector amplitudes and the stator current vector amplitudes. Current fluctuations ( σ _d、 σ _q、 σ Performance indicators such as phase current harmonic distortion (THD) are shown in the figure.

[0073] Depend on Figure 10It can be seen that when the motor operates at 1400 r / min, the average amplitudes of the d-axis and q-axis currents and stator currents of the motor are basically the same under the three methods: PWM-PCC, OVM-PCC, and SPDR-PCC. Compared with PWM-PCC and OVM-PCC, SPDR-PCC shows slightly larger fluctuations in the d-axis current, but smaller fluctuations in the q-axis current, stator current, and xy-axis current. The phase current THDs of PWM-PCC, OVM-PCC, and SPDR-PCC are 5.97%, 6.12%, and 5.30%, respectively. The above results indicate that SPDR-PCC has better steady-state performance under rated operating conditions.

[0074] Depend on Figure 11 It can be seen that: ① When the motor operates at a speed of 1800 r / min and a torque of 10 Nm, the average d-axis currents of the three methods, PWM-PCC, OVM-PCC, and SPDR-PCC, are -2.47 A, -1.89 A, and -1.21 A, respectively. This is because the modulation range of PWM-PCC, OVM-PCC, and SPDR-PCC increases sequentially, thus the required d-axis current decreases sequentially under the same field weakening conditions. ② The average q-axis currents of the three methods, PWM-PCC, OVM-PCC, and SPDR-PCC, are 2.83 A, 2.88 A, and 2.93 A, respectively. However, the average stator current vector amplitudes of PWM-PCC, OVM-PCC, and SPDR-PCC decrease sequentially, to 3.76 A, 3.43 A, and 3.17 A, respectively. Therefore, under the same current limit, the method of this invention can output a larger electromagnetic torque, improving the output capability of the motor torque and speed. ③The fluctuations of d, q, x, and y currents, as well as the fluctuations of stator current vector amplitude, are similar for PWM-PCC, OVM-PCC, and SPDR-PCC.

[0075] The above description of the embodiments is provided to enable those skilled in the art to understand and apply the present invention. Those skilled in the art can readily make various modifications to the above embodiments and apply the general principles described herein to other embodiments without creative effort. Therefore, the present invention is not limited to the above embodiments, and any improvements and modifications made to the present invention by those skilled in the art based on the disclosure thereof should be within the scope of protection of the present invention.

Claims

1. A model predictive control method for an integrated modulated dual three-phase permanent magnet motor, characterized in that, Includes the following steps: (1) Construct a current prediction model with the duty cycle of the six-phase switch signal of the motor inverter as the control quantity to predict the stator current of the dq axis and xy axis at the next moment; (2) Design a full-speed domain current reference value calculation module that combines a field weakening regulator, with the duty cycle of the six-phase switch signal and the electromagnetic torque reference value as input, and output the dq axis current reference value; (3) Use the current prediction error compensator to generate the dq axis current compensation value, which is used to compensate for the dq axis stator current prediction value and dq axis current reference value at the next moment; (4) Establish a cost function that considers the current tracking error of the dq axis and xy axis, and take minimizing the cost function as the optimization objective. Solve to obtain the theoretical optimal value of the duty cycle of the four-phase switch signals A, B, D and E under this objective. (5) Based on the vector synthesis relationship and the theoretical optimal value of the duty cycle of the four-phase switch signals A, B, D, and E, the actual optimal value of the duty cycle of the six-phase switch signals is solved, and then the PWM signal is generated by the double zero-sequence voltage injection pulse width modulation method to drive the inverter of the dual three-phase permanent magnet synchronous motor.

2. The integrated modulation dual three-phase permanent magnet motor model predictive control method according to claim 1, characterized in that, The expression for the current prediction model in step (1) is as follows: in: T s For discrete control periods, , , and Corresponding to kT s The actual values ​​of the stator current along the d-axis, q-axis, x-axis, and y-axis at time t. , , and Corresponding to ( k +1) T s Predicted stator current values ​​along the d-axis, q-axis, x-axis, and y-axis at time points. and They are respectively kT s The rotor electrical angular velocity and rotor electrical angle of the motor at constant time, , , , , and Corresponding to kT s The duty cycle of the A~F phase switch signals of the motor inverter at any given time. This is the stator resistance of the motor; and These are the d-axis inductance and q-axis inductance of the motor, respectively. For the stator leakage inductance of the motor, For permanent magnet flux linkage in electric motors, k It is a natural number.

3. The integrated modulation model predictive control method for dual three-phase permanent magnet motors according to claim 1, characterized in that, The expression for the full-speed domain current reference value calculation module in step (2) is as follows: in: and These are the current reference values ​​for the d-axis and q-axis, respectively. Given the d-axis current value and Set to 0, The field weakening d-axis current generated by the field weakening regulator. This is the reference value for electromagnetic torque. This represents the number of pole pairs of the motor. and These are the d-axis inductance and q-axis inductance of the motor, respectively. This refers to the permanent magnet flux linkage of the motor.

4. The integrated modulation dual three-phase permanent magnet motor model predictive control method according to claim 3, characterized in that: The field weakening regulator uses the maximum value of the six-phase switch signal duty cycle minus 1 as the input to the PI controller. Then, the output of the PI controller is limited and multiplied by the rated current to obtain the field weakening d-axis current. .

5. The integrated modulation model predictive control method for dual three-phase permanent magnet motors according to claim 1, characterized in that, The expression for the current prediction error compensator in step (3) is as follows: in: and These are the current compensation values ​​for the d-axis and q-axis, respectively. and They are respectively kT s The actual values ​​of the stator current along the d-axis and q-axis at time t. and They are ( k +1) T s The predicted stator current values ​​on the d-axis and q-axis at time t, where D() represents the delay function and LPF() represents the low-pass filter function. k It is a natural number.

6. The integrated modulation model predictive control method for dual three-phase permanent magnet motors according to claim 1, characterized in that, The expression for minimizing the cost function in step (4) is as follows: in: J Let cost function be and These are the current compensation values ​​for the d-axis and q-axis, respectively. and They are ( k +2) T s Predicted stator current values ​​along the d-axis and q-axis at time t. and They are ( k +2) T s Predicted stator current values ​​along the x and y axes at time points. and These are the current reference values ​​for the d-axis and q-axis, respectively. and These are the current reference values ​​for the x-axis and y-axis, respectively. λ These are the weighting coefficients. , , , They are ( k +1) T s The theoretical optimal values ​​of the duty cycle of the four-phase switch signals at times A, B, D, and E. k It is a natural number.

7. The integrated modulation model predictive control method for dual three-phase permanent magnet motors according to claim 1, characterized in that, In step (5), the actual optimal value of the duty cycle of the six-phase switch signal is solved using the following relationship: when d A1 ≥0 and d B1 ≥0 and d E1 <0 and d D1 > d E1 In this case, d A = d A1 , d B = d B1 , d C =0, d D = d D1 - d E1 , d E =0, d F =- d E1 ; when d A1 ≥0 and d B1 ≥0 and d D1 ≥0 and d E1 In the case of ≥0, d A = d A1 , d B = d B1 , d C =0, d D = d D1 , d E = d E1 , d F =0; when d A1 <0 and d A1 ≤ d B1 and d D1 ≥0 and d E1 In the case of ≥0, d A =0, d B = d B1 - d A1 , d C =- d A1 , d D = d D1 , d E = d E1 , d F =0; when d A1 <0 and d A1 ≤ d B1 and d D1 <0 and d D1 ≤ d E1 In this case, d A =0, d B = d B1 - d A1 , d C =- d A1 , d D =0, d E = d E1 - d D1 , d F =- d D1 ; when d B1 <0 and d A1 > d B1 and d D1 <0 and d D1 ≤ d E1 In this case, d A = d A1 - d B1 , d B =0, d C =- d B1 , d D =0, d E = d E1 - d D1 , d F =- d D1 ; when d B1 <0 and d A1 > d B1 and d E1 <0 and d D1 > d E1 In this case, d A = d A1 - d B1 , d B =0, d C =- d B1 , d D = d D1 - d E1 , d E =0, d F =- d E1 ; in: d A1 , d B1 , d D1 , d E1 These are the theoretical optimal values ​​for the duty cycles of the four-phase switch signals A, B, D, and E, respectively. d A , d B , d C , d D , d E , d F These are the actual optimal values ​​for the duty cycle of the six-phase switch signals A, B, C, D, E, and F, respectively.

8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that: The processor is used to execute the computer program to implement the integrated modulation dual three-phase permanent magnet motor model predictive control method as described in any one of claims 1 to 7.

9. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by the processor, it implements the integrated modulation dual three-phase permanent magnet motor model predictive control method as described in any one of claims 1 to 7.

Images