Six-phase permanent magnet synchronous motor parameter-free model predictive current control system and method

Abstract

The application discloses a kind of six-phase permanent magnet synchronous motor parameterless model predictive current control system and method, system includes PWM generation module and inverter, PWM generation module is used to output PWM signal control the work of inverter, inverter is used to drive motor;It further includes sampling module, coordinate transformation module, current difference calculation module, hyperlocal model, hyperlocal model parameter estimation module, speed controller, cost function module;Hyperlocal model includes delay compensation module and current prediction module.The application replaces the conventional model by using six-phase permanent magnet synchronous motor hyperlocal model, updates six-phase permanent magnet synchronous motor hyperlocal model parameter by current difference calculation, avoids the dependence of model prediction on motor parameters, can realize high-performance control to six-phase permanent magnet synchronous motor in the case of motor parameter mismatch, effectively solve the problem of strong parameter sensitivity in six-phase permanent magnet synchronous motor model predictive control, improve the robustness of system.

Description

Technical Field

[0001] This invention relates to a synchronous motor control method, and more particularly to a parameterless model predictive current control system and method for a six-phase permanent magnet synchronous motor. Background Technology

[0002] Currently, compared with traditional three-phase motors, multiphase motors have advantages such as low-voltage high-power output and strong fault tolerance, and have broad application prospects in fields such as ship electric propulsion, electric vehicles, and aerospace. Among them, the six-phase permanent magnet synchronous motor with center-point isolation and two sets of windings connected by a 30° phase shift has been widely used due to its special structure, which eliminates six torque ripples. Model predictive control predicts the impact of the current voltage vector on the future performance of the system based on a predictive model. By traversing all possible voltage vectors, it solves the system optimization problem online in a given time domain, and finally applies the selected optimal voltage space vector to the motor to obtain the ideal control effect. Compared with vector control, model predictive control has a simpler control structure, does not require PI tuning, and can obtain fast torque and current responses. At the same time, it can design corresponding optimization objectives for different systems, such as suppressing harmonic currents, improving the quality of load current, and reducing switching frequency. Compared with direct torque control, model predictive control can obtain a more fixed switching frequency and better steady-state performance. Traditional model predictive control employs idealized modeling, neglecting nonlinearity and parameter variations, resulting in a model that heavily relies on the accuracy of system parameters. In practical engineering applications of six-phase permanent magnet synchronous motors, when motor parameters such as resistance and inductance change with temperature and electromagnetic fields, it can lead to motor model mismatch. This can cause torque pulsation or stator winding overcurrent in the motor drive system, reducing the control performance of model predictive control. Summary of the Invention

[0003] This invention provides a parameterless model predictive current control system and method for a six-phase permanent magnet synchronous motor to solve the technical problems existing in the prior art.

[0004] The technical solution adopted by this invention to solve the technical problems existing in the prior art is: a parameterless model predictive current control system for a six-phase permanent magnet synchronous motor, including a PWM generation module and an inverter. The PWM generation module is used to output PWM signals to control the operation of the inverter, and the inverter is used to drive the motor. It also includes a sampling module, a coordinate transformation module, a current difference calculation module, a hyperlocal model, a hyperlocal model parameter estimation module, a speed controller, and a cost function module. The hyperlocal model includes a delay compensation module and a current prediction module.

[0005] The sampling module is used to sample the stator phase current in each control cycle; it outputs the stator phase current signal to the coordinate transformation module.

[0006] The coordinate transformation module is used to convert the stator phase current into d- and q-axis currents in the dq coordinate system; its output d- and q-axis currents are sent to the current difference calculation module, the hyperlocal model parameter estimation module, and the delay compensation module, respectively.

[0007] The current difference calculation module is used to calculate the d-axis current difference and q-axis current difference between adjacent time points; its calculation results are output to the hyperlocal model parameter estimation module.

[0008] The hyperlocal model parameter estimation module is used to estimate the parameters of the hyperlocal model based on the known current difference between the d and q axes and the vector control voltage; its output hyperlocal model parameters are sent to the delay compensation module and the current prediction module, respectively.

[0009] The delay compensation module is used to predict the current value of the motor at time k+1 under the control voltage of the kth control cycle; it outputs the predicted current value at time k+1 to the current prediction module.

[0010] The current prediction module is used to predict the d-axis and q-axis current values ​​at time k+2 under the action of 12 virtual control voltage vectors; its output corresponds to the predicted d-axis and q-axis current values ​​at time k+2 of the 12 virtual control voltage vectors and is sent to the cost function module.

[0011] The speed controller is used to obtain a current reference value based on a given speed; it takes a given speed value as input and outputs the current reference value to the cost function module.

[0012] The cost function module is used to select the optimal vector control voltage based on the error between the current reference value and the current prediction value of each corresponding virtual control voltage vector. It takes the current reference value as input and calculates the error between the current prediction value and the current reference value at time k+2 for each virtual control voltage vector. From the 12 virtual control voltage vectors, it selects the virtual control voltage vector that minimizes the cost function value, which is then used as the control voltage vector V for the k+1 control cycle. opt It will V opt Send to the PWM generation module, which will... opt After being converted into d-axis and q-axis components, they are sent to the hyperlocal model parameter estimation module, delay compensation module, and current prediction module, respectively.

[0013] Furthermore, the mathematical model for the current difference calculation module is as follows:

[0014]

[0015] In the formula:

[0016] Δi d (k) represents the d-axis current difference at time k;

[0017] Δiq (k) represents the q-axis current difference at time k;

[0018] i d (k) represents the d-axis current at time k;

[0019] i q (k) represents the q-axis current at time k;

[0020] i d (k-1) represents the d-axis current at time k-1;

[0021] i q (k-1) represents the q-axis current at time k-1.

[0022] Furthermore, the mathematical model for the delay compensation module is as follows:

[0023]

[0024] In the formula:

[0025] This is the predicted value of the d-axis current at time k+1;

[0026] This is the predicted value of the q-axis current at time k+1;

[0027] F 0_d (k) represents the disturbance parameter of the d-axis corresponding to the hyperlocal model with control period k;

[0028] F 0_q (k) represents the perturbation parameter of the q-axis corresponding to the hyperlocal model with control period k;

[0029] A d (k) represents the gain parameter of the k-axis corresponding to the hyperlocal model with control period k;

[0030] A q (k) represents the gain parameter of the k-axis corresponding to the hyperlocal model with control period;

[0031] V d (k) represents the d-axis control voltage during the k-control period;

[0032] V q (k) represents the q-axis control voltage during the k-control period;

[0033] i d (k) represents the d-axis current at time k;

[0034] i q (k) represents the q-axis current at time k.

[0035] Furthermore, the mathematical model for the current prediction module is as follows:

[0036]

[0037] in:

[0038] V x x = 1 to 12 are 12 virtual control voltage vectors; each is synthesized from two voltage component vectors, which correspond to V respectively. x_L1 and V x_L2 V x_L1 and V x_L2 The directions are the same in the αβ subspace and opposite in the xy subspace; V x The function expression is as follows:

[0039] V x =0.268*V x_L2 +0.732*V x_L1 ;

[0040] In the formula:

[0041] This is the predicted value of the d-axis current at time k+1;

[0042] This is the predicted value of the q-axis current at time k+1;

[0043] This is the predicted value of the d-axis current at time k+2;

[0044] This is the predicted value of the q-axis current at time k+2;

[0045] V dx For virtual control voltage vector V x Voltage component on the d-axis;

[0046] V qx For virtual control voltage vector V x Voltage component on the q-axis.

[0047] Furthermore, the mathematical model for the cost function module is as follows:

[0048]

[0049] In the formula:

[0050] g is the cost function;

[0051] i d * This is the reference value for the d-axis current.

[0052] i q * This is the reference value for the q-axis current.

[0053] This is the predicted value of the d-axis current at time k+2;

[0054] This is the predicted value of the q-axis current at time k+2.

[0055] Furthermore, the hyperlocal model parameter estimation module includes a hyperlocal model d-axis parameter estimation module and a hyperlocal model q-axis parameter estimation module;

[0056] When V d When (k-1) is zero, the mathematical model of the hyperlocal model d-axis parameter estimation module is:

[0057]

[0058] When V d (k-1) is not zero and V d (k-1) is not equal to V d At (k-2), the mathematical model for the hyperlocal model d-axis parameter estimation module is:

[0059]

[0060] When V d (k-1) is not zero and V d (k-1) equals V d When (k-2), the mathematical model of the hyperlocal model d-axis parameter estimation module is:

[0061]

[0062] When V q When (k-1) is zero, the mathematical model of the hyperlocal model q-axis parameter estimation module is:

[0063]

[0064] When V q (k-1) is not zero and V q (k-1) is not equal to V q When (k-2), the mathematical model of the hyperlocal model q-axis parameter estimation module is:

[0065]

[0066] When V q (k-1) is not zero and V q (k-1) equals V q When (k-2), the mathematical model of the hyperlocal model q-axis parameter estimation module is:

[0067]

[0068] In the formula:

[0069] α is the first-order low-pass filter coefficient;

[0070] F 0_d (k-1) represents the disturbance parameter of the d-axis corresponding to the hyperlocal model with control period k-1;

[0071] F 0_q (k-1) represents the disturbance parameter on the q-axis corresponding to the hyperlocal model with a k-1 control period;

[0072] A d (k-1) represents the gain parameter of the d-axis corresponding to the k-1 control period hyperlocal model;

[0073] A q (k-1) represents the gain parameter of the k-1 control period hyperlocal model corresponding to the q-axis;

[0074] Δi d (k) represents the d-axis current difference at time k;

[0075] Δi q (k) represents the q-axis current difference at time k;

[0076] Δi d (k-1) represents the d-axis current difference at time k-1;

[0077] Δi q (k-1) represents the q-axis current difference at time k-1;

[0078] V d (k-1) represents the d-axis control voltage during the k-1 control period;

[0079] V q (k-1) represents the q-axis control voltage during the k-1 control period;

[0080] V d (k-2) represents the q-axis control voltage during the k-2 control period;

[0081] V q (k-2) is the q-axis control voltage during the k-2 control period.

[0082] Furthermore, the speed controller includes a speed PI controller, the mathematical model of which is:

[0083]

[0084] Where: e n =n * -n;

[0085] In the formula:

[0086] iq * This is the reference value for the q-axis current.

[0087] k nP The proportional gain of the speed PI controller;

[0088] k nI The integral gain of the speed PI controller;

[0089] s is a complex variable;

[0090] e n This is the difference between the given rotational speed and the actual rotational speed.

[0091] n is the actual rotational speed;

[0092] n * Given a rotational speed.

[0093] This invention also provides a parameterless model predictive current control method for a six-phase permanent magnet synchronous motor. This method includes a PWM generation module and an inverter. The PWM generation module outputs a PWM signal to control the operation of the inverter, which in turn drives the motor. The method further includes a sampling module, a coordinate transformation module, a current difference calculation module, a hyperlocal model, a hyperlocal model parameter estimation module, a speed controller, and a cost function module. The hyperlocal model includes a delay compensation module and a current prediction module.

[0094] The sampling module is used to sample the stator phase current in each control cycle; and to output the stator phase current signal from the sampling module to the coordinate transformation module.

[0095] The coordinate transformation module is used to convert the stator phase current into d- and q-axis currents in the dq coordinate system; the coordinate transformation module outputs the d- and q-axis currents to the current difference calculation module, the hyperlocal model parameter estimation module, and the delay compensation module, respectively.

[0096] The current difference calculation module is used to calculate the d-axis current difference and q-axis current difference between adjacent time points; the current difference calculation module outputs the calculation results to the hyperlocal model parameter estimation module;

[0097] The hyperlocal model parameter estimation module is used to estimate the parameters of the hyperlocal model based on the known current difference between the d and q axes and the vector control voltage; the hyperlocal model parameter estimation module outputs the parameters of the hyperlocal model to the delay compensation module and the current prediction module respectively;

[0098] The delay compensation module is used to predict the current value of the motor at time k+1 under the control voltage of the kth control cycle; and to make the delay compensation module output the predicted current value at time k+1 to the current prediction module.

[0099] The current prediction module is used to predict the d-axis and q-axis current values ​​at time k+2 under the action of 12 virtual control voltage vectors; and the current prediction module outputs the predicted d-axis and q-axis current values ​​at time k+2 corresponding to the 12 virtual control voltage vectors to the cost function module;

[0100] The speed controller is used to obtain a current reference value based on a given speed; the speed controller takes a given speed value as input and outputs the current reference value to the cost function module.

[0101] The cost function module is used to select the optimal vector control voltage based on the error between the current reference value and the current prediction value of each corresponding virtual control voltage vector. The cost function module takes the current reference value as input and calculates the error between the current prediction value and the current reference value at time k+2 for each virtual control voltage vector. It then selects the virtual control voltage vector that minimizes the cost function value from the 12 virtual control voltage vectors, using it as the control voltage vector V for the (k+1)th control cycle. opt It will V opt Send to the PWM generation module, so that the cost function module will use V opt After being converted into d-axis and q-axis components, they are sent to the hyperlocal model parameter estimation module, delay compensation module, and current prediction module, respectively.

[0102] Furthermore, the method includes the following steps:

[0103] Step 1: Construct mathematical models for the coordinate transformation module, current difference calculation module, delay compensation module, current prediction module, hyperlocal model parameter estimation module, speed controller, and cost function module; assuming the current control cycle is the k-th control cycle, set historical data from control cycles prior to the k-th control cycle, including historical data of motor d-axis and q-axis currents and motor d-axis and q-axis control voltages, and assign initial values ​​to the model parameters of the delay compensation module and the current prediction module;

[0104] Step 2: Obtain the phase current of each phase winding of the motor, the electrical angle of the rotor position, and the motor speed through the sampling module, and obtain the actual current of the d and q axes at each moment through coordinate transformation method;

[0105] Step 3: Using the mathematical model of the current difference calculation module, calculate the d-axis current difference and q-axis current difference between time k and time k-1, and the d-axis current difference and q-axis current difference between time k-1 and time k-2.

[0106] Step 4: Using the mathematical model of the delay compensation module as follows, the predicted current value at time k+1 is obtained from the actual current of the d and q axes at time k and the control voltage of the d and q axes during the kth control period.

[0107]

[0108] In the formula:

[0109] This is the predicted value of the d-axis current at time k+1;

[0110] This is the predicted value of the q-axis current at time k+1;

[0111] F 0_d (k) represents the disturbance parameter of the d-axis corresponding to the hyperlocal model with control period k;

[0112] F 0_q (k) represents the perturbation parameter of the q-axis corresponding to the hyperlocal model with control period k;

[0113] A d (k) represents the gain parameter of the k-axis corresponding to the hyperlocal model with control period k;

[0114] A q (k) represents the gain parameter of the k-axis corresponding to the hyperlocal model with control period;

[0115] V d (k) represents the d-axis control voltage during the k-control period;

[0116] V q (k) represents the q-axis control voltage during the k-control period;

[0117] i d (k) represents the d-axis current at time k;

[0118] i q (k) represents the q-axis current at time k;

[0119] Step 5: Let V x These correspond to 12 virtual control voltage vectors, x = 1 to 12; V dx For V x The voltage component on the d-axis, V qx For virtual control voltage vector V x Voltage component on the q-axis; V x It is synthesized from two voltage vector components, which correspond to V respectively. x_L1 and V x_L2 V x_L1 and V x_L2 The directions are the same in the αβ subspace and opposite in the xy subspace; V x The function expression is as follows:

[0120] V x =0.268*V x_L2 +0.732*V x_L1 ;

[0121] Using the mathematical model of the current prediction module as follows, the predicted current value at time k+1 and the virtual control voltage vector V are used. x The predicted current value at time k+2 is calculated from the voltage component on the q-axis.

[0122]

[0123] In the formula:

[0124] This is the predicted value of the d-axis current at time k+1;

[0125] This is the predicted value of the q-axis current at time k+1;

[0126] This is the predicted value of the d-axis current at time k+2;

[0127] This is the predicted value of the q-axis current at time k+2;

[0128] V dx For virtual control voltage vector V x Voltage component on the d-axis;

[0129] V qx For virtual control voltage vector V x Voltage component on the q-axis;

[0130] Step 6: Using the mathematical model of the local model parameter estimation module, estimate the parameters of the superlocal model based on the known current difference and the corresponding vector control voltage, and update the model parameters of the delay compensation module and the current prediction module.

[0131] Step 7: Speed ​​Controller Setup. Set the speed PI controller, with the d-axis current reference value set to 0. Configure the speed PI controller parameters. Using the speed PI controller, obtain the q-axis current reference value i from the speed setpoint. q * ;

[0132] Step 8: Input the q-axis current reference value i from the speed controller into the cost function module. q * The current prediction value from the current prediction module at time k+2 is used to select the optimal control voltage vector V for the (k+1)th control cycle. opt V opt Send to the PWM generation module; V opt Convert to the corresponding d-axis component V d (k+1) and q-axis component V q (k+1) is sent to the hyperlocal model parameter estimation module, the delay compensation module, and the current prediction module;

[0133] Step 9: Increment k by 1, update the values ​​of the d-axis current difference and the d-axis control voltage, and return to step 2.

[0134] Further, in step 1, the hyperlocal model parameter estimation module is set as the hyperlocal model d-axis parameter estimation module and the hyperlocal model q-axis parameter estimation module; the mathematical models of the hyperlocal model d-axis parameter estimation module and the hyperlocal model q-axis parameter estimation module are constructed as follows:

[0135] When V d When (k-1) is zero, the mathematical model of the hyperlocal model d-axis parameter estimation module is:

[0136]

[0137] When V d (k-1) is not zero and V d (k-1) is not equal to V d At (k-2), the mathematical model for the hyperlocal model d-axis parameter estimation module is:

[0138]

[0139] When V d (k-1) is not zero and V d (k-1) equals V d When (k-2), the mathematical model of the hyperlocal model d-axis parameter estimation module is:

[0140]

[0141] When V q When (k-1) is zero, the mathematical model of the hyperlocal model q-axis parameter estimation module is:

[0142]

[0143] When V q (k-1) is not zero and V q (k-1) is not equal to V q When (k-2), the mathematical model of the hyperlocal model q-axis parameter estimation module is:

[0144]

[0145] When V q (k-1) is not zero and V q (k-1) equals V q When (k-2), the mathematical model of the hyperlocal model q-axis parameter estimation module is:

[0146]

[0147] In the formula:

[0148] α is the first-order low-pass filter coefficient;

[0149] F 0_d (k-1) represents the disturbance parameter of the d-axis corresponding to the hyperlocal model with control period k-1;

[0150] F 0_q (k-1) represents the disturbance parameter on the q-axis corresponding to the hyperlocal model with a k-1 control period;

[0151] A d (k-1) represents the gain parameter of the d-axis corresponding to the k-1 control period hyperlocal model;

[0152] A q (k-1) represents the gain parameter of the k-1 control period hyperlocal model corresponding to the q-axis;

[0153] Δi d (k) represents the d-axis current difference at time k;

[0154] Δi q (k) represents the q-axis current difference at time k;

[0155] Δi d (k-1) represents the d-axis current difference at time k-1;

[0156] Δi q (k-1) represents the q-axis current difference at time k-1;

[0157] V d (k-1) represents the d-axis control voltage during the k-1 control period;

[0158] V q (k-1) represents the q-axis control voltage during the k-1 control period;

[0159] V d (k-2) represents the q-axis control voltage during the k-2 control period;

[0160] V q (k-2) is the q-axis control voltage during the k-2 control period.

[0161] The advantages and positive effects of this invention are as follows: The present invention proposes a parameterless model predictive current control system and method for a six-phase permanent magnet synchronous motor. By using a hyperlocal model of the six-phase permanent magnet synchronous motor to replace the traditional model of the permanent magnet synchronous motor, and by calculating and updating the parameters of the hyperlocal model through the current difference, the dependence of model predictive control on the accuracy of motor parameters is eliminated, effectively solving the parameter sensitivity problem of model predictive control and greatly improving the robustness of the system. Attached Figure Description

[0162] Figure 1 This is a block diagram of a parameterless model predictive current control system for a six-phase permanent magnet synchronous motor according to the present invention;

[0163] Figure 2 The distribution diagram of the 12 virtual control voltage vectors and the zero voltage vector;

[0164] Figure 3 The distribution diagram of voltage vector components in the αβ coordinate system decomposed into 12 virtual control voltage vectors;

[0165] Figure 4 The distribution diagram of voltage vector components in the xy coordinate system decomposed into 12 virtual control voltage vectors;

[0166] Figure 5 The waveform diagram of the parameters of the hyperlocal model F0(k) after Euler forward discretization of a six-phase permanent magnet synchronous motor is shown.

[0167] Figure 6 The waveform diagram of the parameters of the hyperlocal model A(k) after Euler forward discretization of a six-phase permanent magnet synchronous motor is shown.

[0168] Figure 7 The image shows the motor current waveform using the parameterless model predictive current control method for a six-phase permanent magnet synchronous motor according to the present invention.

[0169] Figure 8 The image shows the torque waveform of a six-phase permanent magnet synchronous motor using a parameterless model predictive current control method according to the present invention.

[0170] Figure 9 The image shows the motor speed waveform using the parameterless model predictive current control method for a six-phase permanent magnet synchronous motor according to the present invention.

[0171] Figure 10 This is a diagram showing the correspondence between the control cycle and time in this invention.

[0172] In the picture:

[0173] i abc (k) represents the currents of phases a, b, and c at time k;

[0174] i uvw (k) represents the phase currents u, v, and w at time k;

[0175] s abc (k) represents the switching signals of the a, b, and c phase bridge arms at time k;

[0176] s uvw (k) represents the switching signals of phase u, v, and w of the bridge arm at time k;

[0177] V optThe optimal voltage vector;

[0178] i a Let a be the phase current;

[0179] i u The current is the u-phase current.

[0180] Te For output torque;

[0181] θ e The position is an electrical angle;

[0182] n is the motor speed;

[0183] V 0～ V 12 It consists of a zero voltage vector and 12 virtual voltage vectors;

[0184] This is the predicted value of the d-axis current at time k+1;

[0185] This is the predicted value of the q-axis current at time k+1;

[0186] F0(k) characterizes the total disturbance parameter of the hyperlocal model of the six-phase permanent magnet synchronous motor in the kth control cycle;

[0187] A(k) characterizes the control gain parameter of the hyperlocal model of a six-phase permanent magnet synchronous motor with a control period of k.

[0188] F 0_d For the perturbation parameters of the hyperlocal model corresponding to the d-axis;

[0189] F 0_q These are the perturbation parameters for the q-axis corresponding to the hyperlocal model;

[0190] A d This represents the gain parameter along the d-axis of the hyperlocal model.

[0191] A q This represents the gain parameter along the q-axis corresponding to the hyperlocal model;

[0192] V d (k) represents the d-axis control voltage during the k-control period;

[0193] V q (k) represents the q-axis control voltage during the k-control period;

[0194] i d (k) represents the d-axis current at time k;

[0195] i q (k) represents the q-axis current at time k;

[0196] Δi d(k) represents the d-axis current difference at time k;

[0197] Δi q (k) represents the q-axis current difference at time k;

[0198] i d (k) represents the d-axis current at time k;

[0199] i q (k) represents the q-axis current at time k;

[0200] i d (k-1) represents the d-axis current at time k-1;

[0201] i q (k-1) represents the q-axis current at time k-1;

[0202] This is the predicted value of the d-axis current at time k+1;

[0203] This is the predicted value of the q-axis current at time k+1;

[0204] This is the predicted value of the d-axis current at time k+2;

[0205] This is the predicted value of the q-axis current at time k+2;

[0206] i d * This is the reference value for the d-axis current.

[0207] i q * This is the reference value for the q-axis current.

[0208] k nP The proportional gain of the speed PI controller;

[0209] k nI The integral gain of the speed PI controller;

[0210] s is a complex variable;

[0211] e n This is the difference between the given rotational speed and the actual rotational speed.

[0212] n is the actual rotational speed;

[0213] n * Given a rotational speed. Detailed Implementation

[0214] To further understand the invention's content, features, and effects, the following embodiments are provided, along with detailed descriptions in conjunction with the accompanying drawings:

[0215] Please see Figures 1 to 10 A parameterless model predictive current control system for a six-phase permanent magnet synchronous motor includes a PWM generation module and an inverter. The PWM generation module is used to output PWM signals to control the operation of the inverter, and the inverter is used to drive the motor. It also includes a sampling module, a coordinate transformation module, a current difference calculation module, a hyperlocal model, a hyperlocal model parameter estimation module, a speed controller, and a cost function module. The hyperlocal model includes a delay compensation module and a current prediction module.

[0216] The sampling module is used to sample the stator phase current in each control cycle; it outputs the stator phase current signal to the coordinate transformation module.

[0217] The coordinate transformation module is used to convert the stator phase current into d- and q-axis currents in the dq coordinate system; its output d- and q-axis currents are sent to the current difference calculation module, the hyperlocal model parameter estimation module, and the delay compensation module, respectively.

[0218] The current difference calculation module is used to calculate the d-axis current difference and q-axis current difference between adjacent time points; its calculation results are output to the hyperlocal model parameter estimation module.

[0219] The hyperlocal model parameter estimation module is used to estimate the parameters of the hyperlocal model based on the known current difference between the d and q axes and the vector control voltage; its output hyperlocal model parameters are sent to the delay compensation module and the current prediction module, respectively.

[0220] The delay compensation module is used to predict the current value of the motor at time k+1 under the control voltage of the kth control cycle; it outputs the predicted current value at time k+1 to the current prediction module.

[0221] The current prediction module is used to predict the d-axis and q-axis current values ​​at time k+2 under the action of 12 virtual control voltage vectors; its output corresponds to the predicted d-axis and q-axis current values ​​at time k+2 of the 12 virtual control voltage vectors and is sent to the cost function module.

[0222] The speed controller is used to obtain a current reference value based on a given speed; it takes a given speed value as input and outputs the current reference value to the cost function module.

[0223] The cost function module is used to select the optimal vector control voltage based on the error between the current reference value and the current prediction value of each corresponding virtual control voltage vector. It takes the current reference value as input and calculates the error between the current prediction value and the current reference value at time k+2 for each virtual control voltage vector. From the 12 virtual control voltage vectors, it selects the virtual control voltage vector that minimizes the cost function value, which is then used as the control voltage vector V for the k+1 control cycle. opt It will V optSend to the PWM generation module, which will... opt After being converted into d-axis and q-axis components, they are sent to the hyperlocal model parameter estimation module, delay compensation module, and current prediction module, respectively.

[0224] Preferably, the mathematical model for the current difference calculation module can be:

[0225]

[0226] In the formula:

[0227] Δi d (k) represents the d-axis current difference at time k;

[0228] Δi q (k) represents the q-axis current difference at time k;

[0229] i d (k) represents the d-axis current at time k;

[0230] i q (k) represents the q-axis current at time k;

[0231] i d (k-1) represents the d-axis current at time k-1;

[0232] i q (k-1) represents the q-axis current at time k-1.

[0233] Preferably, the mathematical model of the delay compensation module can be:

[0234]

[0235] In the formula:

[0236] This is the predicted value of the d-axis current at time k+1;

[0237] This is the predicted value of the q-axis current at time k+1;

[0238] F 0d (k) represents the disturbance parameter of the d-axis corresponding to the hyperlocal model with control period k;

[0239] F 0_q (k) represents the perturbation parameter of the q-axis corresponding to the hyperlocal model with control period k;

[0240] A d (k) represents the gain parameter of the k-axis corresponding to the hyperlocal model with control period k;

[0241] A q (k) represents the gain parameter of the k-axis corresponding to the hyperlocal model with control period;

[0242] V d (k) represents the d-axis control voltage during the k-control period;

[0243] V q (k) represents the q-axis control voltage during the k-control period;

[0244] i d (k) represents the d-axis current at time k;

[0245] i q (k) represents the q-axis current at time k.

[0246] Preferably, the mathematical model of the current prediction module can be:

[0247]

[0248] in:

[0249] V x x = 1 to 12 are 12 virtual control voltage vectors; each is synthesized from two voltage component vectors, which correspond to V respectively. x_L1 and V x_L2 V x_L1 and V x_L2 The directions are the same in the αβ subspace and opposite in the xy subspace; V x The function expression is as follows:

[0250] V x =0.268*V x_L2 +0.732*V x_L1 ;

[0251] In the formula:

[0252] This is the predicted value of the d-axis current at time k+1;

[0253] This is the predicted value of the q-axis current at time k+1;

[0254] This is the predicted value of the d-axis current at time k+2;

[0255] This is the predicted value of the q-axis current at time k+2;

[0256] V dx For virtual control voltage vector V x Voltage component on the d-axis;

[0257] V qx For virtual control voltage vector V x Voltage component on the q-axis.

[0258] Preferably, the mathematical model of the cost function module can be:

[0259]

[0260] In the formula:

[0261] g is the cost function;

[0262] i d * This is the reference value for the d-axis current.

[0263] i q * This is the reference value for the q-axis current.

[0264] This is the predicted value of the d-axis current at time k+2;

[0265] This is the predicted value of the q-axis current at time k+2.

[0266] Preferably, the hyperlocal model parameter estimation module may include a hyperlocal model d-axis parameter estimation module and a hyperlocal model q-axis parameter estimation module;

[0267] When V d When (k-1) is zero, the mathematical model of the hyperlocal model d-axis parameter estimation module can be:

[0268]

[0269] When V d (k-1) is not zero and V d (k-1) is not equal to V d When (k-2), the mathematical model for the hyperlocal model d-axis parameter estimation module can be:

[0270]

[0271] When V d (k-1) is not zero and V d (k-1) equals V d When (k-2), the mathematical model of the hyperlocal model d-axis parameter estimation module can be:

[0272]

[0273] When V q When (k-1) is zero, the mathematical model of the hyperlocal model q-axis parameter estimation module can be:

[0274]

[0275] When V q (k-1) is not zero and Vq (k-1) is not equal to V q When (k-2), the mathematical model of the hyperlocal model q-axis parameter estimation module can be:

[0276]

[0277] When V q (k-1) is not zero and V q (k-1) equals V q When (k-2), the mathematical model of the hyperlocal model q-axis parameter estimation module can be:

[0278]

[0279] In the formula:

[0280] α is the first-order low-pass filter coefficient;

[0281] F 0_d (k-1) represents the disturbance parameter of the d-axis corresponding to the hyperlocal model with control period k-1;

[0282] F 0_q (k-1) represents the disturbance parameter on the q-axis corresponding to the hyperlocal model with a k-1 control period;

[0283] A d (k-1) represents the gain parameter of the d-axis corresponding to the k-1 control period hyperlocal model;

[0284] A q (k-1) represents the gain parameter of the k-1 control period hyperlocal model corresponding to the q-axis;

[0285] Δi d (k) represents the d-axis current difference at time k;

[0286] Δi q (k) represents the q-axis current difference at time k;

[0287] Δi d (k-1) represents the d-axis current difference at time k-1;

[0288] Δi q (k-1) represents the q-axis current difference at time k-1;

[0289] V d (k-1) represents the d-axis control voltage during the k-1 control period;

[0290] V q (k-1) represents the q-axis control voltage during the k-1 control period;

[0291] V d (k-2) represents the q-axis control voltage during the k-2 control period;

[0292] V q (k-2) is the q-axis control voltage during the k-2 control period.

[0293] Preferably, the speed controller includes a speed PI controller, and the mathematical model of the speed PI controller can be:

[0294]

[0295] Where: e n =n * -n;

[0296] In the formula:

[0297] i q * This is the reference value for the q-axis current.

[0298] k nP The proportional gain of the speed PI controller;

[0299] k nI The integral gain of the speed PI controller;

[0300] s is a complex variable;

[0301] e n This is the difference between the given rotational speed and the actual rotational speed.

[0302] n is the actual rotational speed;

[0303] n * Given a rotational speed.

[0304] This invention also provides a parameterless model predictive current control method for a six-phase permanent magnet synchronous motor. This method includes a PWM generation module and an inverter. The PWM generation module outputs a PWM signal to control the operation of the inverter, which in turn drives the motor. The method further includes a sampling module, a coordinate transformation module, a current difference calculation module, a hyperlocal model, a hyperlocal model parameter estimation module, a speed controller, and a cost function module. The hyperlocal model includes a delay compensation module and a current prediction module.

[0305] The sampling module is used to sample the stator phase current in each control cycle; and to output the stator phase current signal from the sampling module to the coordinate transformation module.

[0306] The coordinate transformation module is used to convert the stator phase current into d- and q-axis currents in the dq coordinate system; the coordinate transformation module outputs the d- and q-axis currents to the current difference calculation module, the hyperlocal model parameter estimation module, and the delay compensation module, respectively.

[0307] The current difference calculation module is used to calculate the d-axis current difference and q-axis current difference between adjacent time points; the current difference calculation module outputs the calculation results to the hyperlocal model parameter estimation module;

[0308] The hyperlocal model parameter estimation module is used to estimate the parameters of the hyperlocal model based on the known current difference between the d and q axes and the vector control voltage; the hyperlocal model parameter estimation module outputs the parameters of the hyperlocal model to the delay compensation module and the current prediction module respectively;

[0309] The delay compensation module is used to predict the current value of the motor at time k+1 under the control voltage of the kth control cycle; and to make the delay compensation module output the predicted current value at time k+1 to the current prediction module.

[0310] The current prediction module is used to predict the d-axis and q-axis current values ​​at time k+2 under the action of 12 virtual control voltage vectors; and the current prediction module outputs the predicted d-axis and q-axis current values ​​at time k+2 corresponding to the 12 virtual control voltage vectors to the cost function module;

[0311] The speed controller is used to obtain a current reference value based on a given speed; the speed controller takes a given speed value as input and outputs the current reference value to the cost function module.

[0312] The cost function module is used to select the optimal vector control voltage based on the error between the current reference value and the current prediction value of each corresponding virtual control voltage vector. The cost function module takes the current reference value as input and calculates the error between the current prediction value and the current reference value at time k+2 for each virtual control voltage vector. It then selects the virtual control voltage vector that minimizes the cost function value from the 12 virtual control voltage vectors, using it as the control voltage vector V for the (k+1)th control cycle. opt It will V opt Send to the PWM generation module, so that the cost function module will use V opt After being converted into d-axis and q-axis components, they are sent to the hyperlocal model parameter estimation module, delay compensation module, and current prediction module, respectively.

[0313] Preferably, the method includes the following steps:

[0314] Step 1: Mathematical models for the coordinate transformation module, current difference calculation module, delay compensation module, current prediction module, hyperlocal model parameter estimation module, speed controller, and cost function module can be constructed. The current control cycle can be set as the k-th control cycle, and historical data of control cycles before the k-th control cycle can be set, including historical data of motor d-axis and q-axis current and motor d-axis and q-axis control voltage. Initial values ​​are assigned to the model parameters of the delay compensation module and the current prediction module.

[0315] The following is a basic model for establishing a hyperlocal model of a six-phase permanent magnet synchronous motor after Euler forward discretization:

[0316]

[0317] In the formula, F0(k)=[F 0_d (k)F 0_q (k)] and A(k)=[A d (k)A q [(k)] represents the parameters of the hyperlocal model of a six-phase permanent magnet synchronous motor.

[0318] in:

[0319] F0(k) characterizes the total disturbance parameter of the hyperlocal model of the six-phase permanent magnet synchronous motor in the kth control cycle;

[0320] A(k) characterizes the control gain parameter of the hyperlocal model of a six-phase permanent magnet synchronous motor with a control period of k.

[0321] In the formula:

[0322] This is the predicted value of the d-axis current at time k+1;

[0323] This is the predicted value of the q-axis current at time k+1;

[0324] F 0_d (k) represents the disturbance parameter of the d-axis corresponding to the hyperlocal model with control period k;

[0325] F 0_q (k) represents the perturbation parameter of the q-axis corresponding to the hyperlocal model with control period k;

[0326] A d (k) represents the gain parameter of the k-axis corresponding to the hyperlocal model with control period k;

[0327] A q (k) represents the gain parameter of the k-axis corresponding to the hyperlocal model with control period;

[0328] V d (k) represents the d-axis control voltage during the k-control period;

[0329] V q (k) represents the q-axis control voltage during the k-control period;

[0330] i d (k) represents the d-axis current at time k;

[0331] i q (k) represents the q-axis current at time k.

[0332] We can assume that the d-axis and q-axis components of the control voltage vector in the (k-1)th control cycle correspond to V. d (k-1), V q (k-1); it can be assumed that the d-axis and q-axis components of the control voltage vector in the (k-2)th control cycle correspond to V. d (k-2), V q (k-2); V can be preset d (k-1), V q (k-1), V d (k-2), V q (k-2), F 0_d (k), F 0_q (k), A d (k), A q The initial value of (k).

[0333] Based on this hyperlocal model, mathematical models for the delay compensation module and the current prediction module are constructed. The model parameters of the mathematical models for both the delay compensation module and the current prediction module can be the basic model parameters of the hyperlocal model.

[0334] Step 2: The phase current of each phase winding of the motor, the electrical angle of the rotor position and the motor speed can be obtained through the sampling module, and the actual current of the d and q axes at each moment can be obtained through coordinate transformation.

[0335] Please see Figure 10 The relationship between the sampling period and the control period is as follows: the period between time k-1 and time k is the k-1 control period, the period between time k and time k+1 is the k control period, time k-1 corresponds to the k-1 sampling time, time k corresponds to the k sampling time, and so on.

[0336] At time k, the current i of the six-phase permanent magnet synchronous motor is obtained through a current sensor. a (k), i b (k), i c (k), i u (k), i v (k), i w (k), the rotor position electrical angle θ is obtained through the encoder. e Given the motor speed n, the dq-axis current i at time k is obtained through coordinate transformation. d (k) and i q (k), whose coordinate transformation matrix is:

[0337]

[0338] In the formula:

[0339] i a (k) represents the sampled value of the phase current at time k;

[0340] i b (k) represents the sampled value of the phase b current at time k;

[0341] i c (k) represents the sampled value of the c-phase current at time k;

[0342] i u (k) represents the sampled value of the u-phase current at time k;

[0343] i v (k) represents the sampled value of the phase v current at time k;

[0344] i w (k) represents the sampled value of the phase current w at time k;

[0345] i d (k) represents the d-axis current at time k;

[0346] i q (k) represents the q-axis current at time k.

[0347] The sampling module can obtain the phase currents of each phase winding at times k-2, k-1, and k, as well as the rotor position electrical angles and motor speeds at times k-2, k-1, and k. The actual d-axis and q-axis currents i at time k can be obtained through coordinate transformation. d (k) and i q (k) and the actual currents i on the d and q axes at time k-1 d (k-1), i q (k-1).

[0348] Step 3: The mathematical model of the current difference calculation module can be used to calculate the d-axis current difference and q-axis current difference between time k and time k-1, and the d-axis current difference and q-axis current difference between time k-1 and time k-2.

[0349] The mathematical model of the following current difference calculation module can be used to calculate the current difference Δi between the d and q axes at time k. d (k), Δi q (k), and the current difference Δi between the d and q axes at time k-1. d (k-1), Δi q (k-1):

[0350]

[0351] In the formula:

[0352] Δi d (k) represents the d-axis current difference at time k;

[0353] Δi q(k) represents the q-axis current difference at time k;

[0354] i d (k) represents the d-axis current at time k;

[0355] i q (k) represents the q-axis current at time k;

[0356] i d (k-1) represents the d-axis current at time k-1;

[0357] i q (k-1) represents the q-axis current at time k-1.

[0358] Step 4: The mathematical model of the delay compensation module adopts the basic model of the hyperlocal model. The following mathematical model of the delay compensation module can be used to obtain the predicted current value at time k+1 from the actual current of the d and q axes at time k and the control voltage of the d and q axes in the kth control period.

[0359]

[0360] In the formula:

[0361] This is the predicted value of the d-axis current at time k+1;

[0362] This is the predicted value of the q-axis current at time k+1;

[0363] F 0_d (k) represents the disturbance parameter of the d-axis corresponding to the hyperlocal model with control period k;

[0364] F 0_q (k) represents the perturbation parameter of the q-axis corresponding to the hyperlocal model with control period k;

[0365] A d (k) represents the gain parameter of the k-axis corresponding to the hyperlocal model with control period k;

[0366] A q (k) represents the gain parameter of the k-axis corresponding to the hyperlocal model with control period;

[0367] V d (k) represents the d-axis control voltage during the k-control period;

[0368] V q (k) represents the q-axis control voltage during the k-control period;

[0369] i d (k) represents the d-axis current at time k;

[0370] i q (k) represents the q-axis current at time k.

[0371] Step 5: V can be set x These correspond to 12 virtual control voltage vectors, x = 1 to 12; V dx For virtual control voltage vector V x The voltage component on the d-axis, V qx For virtual control voltage vector V x Voltage component on the q-axis; V x It is synthesized from two voltage vector components, which correspond to V respectively. x_L1 and V x_L2 V x_L1 and V x_L2 The directions are the same in the αβ subspace and opposite in the xy subspace. Wherein:

[0372] αβ subspace: The fundamental component and the 12k±1 (k=l,2,3,…) harmonic components in the motor variables are mapped onto this subspace and participate in the electromechanical energy conversion of the motor.

[0373] The xy subspace: 6k±1 (k=l,2,3,…) harmonic components are mapped onto this subspace and do not participate in the electromechanical energy conversion of the motor.

[0374] V x The function expression is as follows:

[0375] V x =0.268*V x_L2 +0.732*V x_L1 ;

[0376] The mathematical model of the current prediction module has been improved based on the hyperlocal model, and a virtual control voltage vector is used to replace the control voltage.

[0377] The following mathematical model of the current prediction module can be used, based on the predicted current value at time k+1 and the virtual control voltage vector V. x The predicted current value at time k+2 is calculated from the voltage component on the q-axis.

[0378]

[0379] In the formula:

[0380] This is the predicted value of the d-axis current at time k+1;

[0381] This is the predicted value of the q-axis current at time k+1;

[0382] This is the predicted value of the d-axis current at time k+2;

[0383] This is the predicted value of the q-axis current at time k+2;

[0384] V dx For virtual control voltage vector V x Voltage component on the d-axis;

[0385] V qx For virtual control voltage vector V x Voltage component on the q-axis.

[0386] The current prediction module calculates the predicted current value at time k+2 under the action of the 12 virtual control voltage vectors and the zero voltage vector. and

[0387] Among them, 12 virtual control voltage vectors V x The distribution graph of (x=1~12) is as follows Figure 2 As shown, each virtual control voltage vector is synthesized from two voltage vectors. Furthermore, the two voltage vectors that synthesize the virtual control voltage vector are in opposite directions in the xy subspace. By allocating the application time of the two voltage vectors, their effects in the xy subspace are canceled out. See also... Figures 2 to 4 For example, for the virtual vector V1(4-4,6-5), V x_L1 =V 4-4 V x_L2 =V 6-5 Among them, V 4-4 and V 6-5 This is the basic voltage vector that the inverter can output; V1(4-4,6-5) is V 4-4 With V 6-5 The synthesized voltage vector.

[0388] Step 6: The mathematical model of the local model parameter estimation module can be used to estimate the parameters of the superlocal model based on the known current difference and the corresponding vector control voltage, and the model parameters of the delay compensation module and the current prediction module can be updated.

[0389] Step 7: The speed controller can be configured as a PI controller. The d-axis current reference value can be set to 0. The parameters of the PI controller can be set, and the q-axis current reference value i can be obtained from the speed setpoint using the PI controller. q * .

[0390] The following mathematical model for a speed PI controller can be constructed:

[0391]

[0392] Where: e n =n *-n;

[0393] In the formula:

[0394] i q * This is the reference value for the q-axis current.

[0395] k nP The proportional gain of the speed PI controller;

[0396] k nI The integral gain of the speed PI controller;

[0397] s is a complex variable;

[0398] e n This is the difference between the given rotational speed and the actual rotational speed.

[0399] n is the actual rotational speed;

[0400] n * Given a rotational speed.

[0401] Step 8: The cost function module can input the q-axis current reference value i from the speed controller. q * The system calculates the current prediction value from the current prediction module at time k+2, and then calculates the error between the current prediction value and the current reference value at time k+2 for each virtual control voltage vector. Finally, it selects the optimal control voltage vector V for the (k+1)th control cycle. opt .

[0402] Calculate the following errors separately: the error between the predicted current value and the reference current value at time k+2 corresponding to the first virtual control voltage vector; and the error between the predicted current value and the reference current value at time k+2 corresponding to the 12th virtual control voltage vector. Correspond the calculated errors to the virtual control voltage vectors, and select the virtual control voltage vector that minimizes the cost function value from the 12 virtual control voltage vectors as the optimal control voltage vector V for the (k+1)th control cycle. opt V opt The signal is sent to the PWM generation module, which generates the PWM signal to drive the power switches of the six-phase inverter. V opt Convert to the corresponding d-axis component V d (k+1) and q-axis component V q (k+1) is sent to the hyperlocal model parameter estimation module, the delay compensation module, and the current prediction module.

[0403] Preferably, the mathematical model for constructing the cost function module is as follows:

[0404]

[0405]

[0406] V opt This represents the optimal control voltage vector for the (k+1)th control cycle.

[0407] g is the cost function;

[0408] i d * This is the reference value for the d-axis current.

[0409] i q * This is the reference value for the q-axis current.

[0410] V d (k+1) represents the d-axis control voltage during the k+1 control period;

[0411] V q (k+1) represents the q-axis control voltage during the k+1 control period;

[0412] This is the predicted value of the d-axis current at time k+2;

[0413] This is the predicted value of the q-axis current at time k+2.

[0414] Step 9: Increment k by 1, update the values ​​of the d-axis current difference and the d-axis control voltage, and return to step 2.

[0415] Preferably, in step 1, the hyperlocal model parameter estimation module is set as a hyperlocal model d-axis parameter estimation module and a hyperlocal model q-axis parameter estimation module; the mathematical models of the hyperlocal model d-axis parameter estimation module and the hyperlocal model q-axis parameter estimation module are constructed as follows:

[0416] When V d When (k-1) is zero, the mathematical model of the hyperlocal model d-axis parameter estimation module is:

[0417]

[0418] When V d (k-1) is not zero and V d (k-1) is not equal to V d At (k-2), the mathematical model for the hyperlocal model d-axis parameter estimation module is:

[0419]

[0420] When V d (k-1) is not zero and V d (k-1) equals V dWhen (k-2), the mathematical model of the hyperlocal model d-axis parameter estimation module is:

[0421]

[0422] When V q When (k-1) is zero, the mathematical model of the hyperlocal model q-axis parameter estimation module is:

[0423]

[0424] When V q (k-1) is not zero and V q (k-1) is not equal to V q When (k-2), the mathematical model of the hyperlocal model q-axis parameter estimation module is:

[0425]

[0426] When V q (k-1) is not zero and V q (k-1) equals V q When (k-2), the mathematical model of the hyperlocal model q-axis parameter estimation module is:

[0427]

[0428] In the formula:

[0429] α is the first-order low-pass filter coefficient;

[0430] F 0_d (k-1) represents the disturbance parameter of the d-axis corresponding to the hyperlocal model with control period k-1;

[0431] F 0_q (k-1) represents the disturbance parameter on the q-axis corresponding to the hyperlocal model with a k-1 control period;

[0432] A d (k-1) represents the gain parameter of the d-axis corresponding to the k-1 control period hyperlocal model;

[0433] A q (k-1) represents the gain parameter of the k-1 control period hyperlocal model corresponding to the q-axis;

[0434] Δi d (k) represents the d-axis current difference at time k;

[0435] Δi q (k) represents the q-axis current difference at time k;

[0436] Δi d (k-1) represents the d-axis current difference at time k-1;

[0437] Δi q (k-1) represents the q-axis current difference at time k-1;

[0438] V d (k-1) represents the d-axis control voltage during the k-1 control period;

[0439] V q (k-1) represents the q-axis control voltage during the k-1 control period;

[0440] V d (k-2) represents the q-axis control voltage during the k-2 control period;

[0441] V q (k-2) is the q-axis control voltage during the k-2 control period.

[0442] Based on the known Δi d (k), Δi q (k), Δi d (k-1), Δi q (k-1), V d (k-1), V q (k-1), V d (k-2), V q (k-2) updates the parameters of the hyperlocal model to obtain the updated parameters A(k) and F0(k) of the hyperlocal model in the kth control cycle.

[0443] The aforementioned PWM generation module, inverter, sampling module, coordinate transformation module, current difference calculation module, hyperlocal model, hyperlocal model parameter estimation module, speed controller, cost function module, delay compensation module, current prediction module, and speed PI controller can all use existing components and functional modules, or be constructed using existing components and functional modules and conventional technical means.

[0444] To verify the effectiveness of this invention, a parameterless model predictive current control method for a six-phase permanent magnet synchronous motor proposed in this invention was simulated in the MATLAB / Simulink environment. The simulation was performed with a given speed of 500 r / min and a given torque of 5 N·m.

[0445] like Figure 5 , Figure 6 The figure shows the waveform of the hyperlocal model parameters after Euler forward discretization of a six-phase permanent magnet synchronous motor. It can be seen from the figure that the hyperlocal model parameters F... 0_d F 0_d A d A qThe system stabilizes within 0.06 seconds after startup, with fluctuations not exceeding 0.1. Simulation results demonstrate that the proposed method for calculating and updating parameters of the hyperlocal model after Euler forward discretization of the six-phase permanent magnet synchronous motor exhibits excellent dynamic and steady-state performance.

[0446] like Figures 7 to 9 The figure shows the current, speed, and torque waveforms of a six-phase permanent magnet synchronous motor. The motor speed n reaches the given value of 500 r / min in 0.02 s. The motor output torque is the maximum allowable output torque during the acceleration phase, and quickly becomes 5 N·m after the motor speed reaches the given value. The motor phase current is sinusoidal. Simulation results show that the parameterless model predictive current control method for a six-phase permanent magnet synchronous motor proposed in this invention has good dynamic and steady-state performance.

[0447] The embodiments described above are only used to illustrate the technical ideas and features of the present invention. Their purpose is to enable those skilled in the art to understand the content of the present invention and implement it accordingly. The patent scope of the present invention should not be limited by these embodiments. That is, any equivalent changes or modifications made in accordance with the spirit disclosed in the present invention still fall within the patent scope of the present invention.

Claims

1. A parameter-free model predictive current control system for a six-phase permanent magnet synchronous motor, comprising a PWM generation module and an inverter, the PWM generation module is used to output PWM signals to control the operation of the inverter, and the inverter is used to drive the motor; characterized in that, The method further comprises a sampling module, a coordinate transformation module, a current difference calculation module, a hyper-local model, a hyper-local model parameter estimation module, a speed controller, and a cost function module; the hyper-local model comprises a time delay compensation module and a current prediction module; The sampling module is configured to sample stator phase currents of each control period; the sampling module outputs stator phase current signals to the coordinate transformation module; The coordinate transformation module is configured to convert the stator phase currents into d-axis and q-axis currents in a dq coordinate system; the coordinate transformation module outputs the d-axis and q-axis currents to the current difference calculation module, the hyper-local model parameter estimation module, and the time delay compensation module, respectively; The current difference calculation module is configured to calculate d-axis and q-axis current differences between adjacent time instants; The current difference calculation module outputs calculation results to the hyper-local model parameter estimation module; The hyper-local model parameter estimation module is configured to estimate parameters of the hyper-local model according to known d-axis and q-axis current differences and vector control voltages; the hyper-local model parameter estimation module outputs the parameters of the hyper-local model to the time delay compensation module and the current prediction module, respectively; The time delay compensation module is configured to predict current values of the motor at a k+1 time instant under the control of a k control period control voltage; the time delay compensation module outputs the current prediction values at the k+1 time instant to the current prediction module; The current prediction module is configured to predict d-axis and q-axis current values at a k+2 time instant under the control of 12 virtual control voltage vectors, respectively; the current prediction module outputs the d-axis and q-axis current prediction values corresponding to the 12 virtual control voltage vectors at the k+2 time instant to the cost function module; The speed controller is configured to obtain current reference values from a given speed; The speed controller inputs the given speed value and outputs the current reference values to the cost function module; The cost function module is configured to select an optimal vector control voltage according to errors between the current reference values and the current prediction values corresponding to the virtual control voltage vectors; The input current reference value is inputted, and the error between the current prediction value corresponding to each virtual control voltage vector at the k+2 moment and the current reference value is calculated respectively, and from the 12 virtual control voltage vectors, the virtual control voltage vector making the cost function value minimum is selected as the control voltage vector V opt of the k+1 control period opt , which is sent to the PWM generation module opt , which converts V opt into d-axis component and q-axis component and sends them to the hyperlocal model parameter estimation module, the delay compensation module and the current prediction module respectively.

2. The six-phase permanent magnet synchronous motor parameter-less model predictive current control system according to claim 1, characterized in that, The mathematical model of the current difference calculation module is as follows: In the formula, Δi d (k) is the d-axis current difference at time k; Δi q (k) is the q-axis current difference at time k; i d (k) is the d-axis current at time k; i q (k) is the q-axis current at time k; i d (k-1) is the d-axis current at time k-1; i q (k-1) is the q-axis current at time k-1.

3. The six-phase permanent magnet synchronous motor parameter-less model predictive current control system according to claim 1, characterized in that, The mathematical model of the time delay compensation module is as follows: In the formula, is the d-axis current prediction value for the k+1 instant; is the q-axis current prediction value for the k+1 instant; F 0_d (k) is a disturbance parameter of the d-axis corresponding to the k control period super-local model; F 0_q (k) is the disturbance parameter of the q-axis corresponding to the k control period super-local model; A d (k) is a gain parameter of the d-axis corresponding to the k control period hyperlocal model; A q (k) is a gain parameter of the q-axis corresponding to the k control period hyperlocal model; V d (k) is a k control period d-axis control voltage; V q (k) is a k control period q-axis control voltage; i d (k) is the d-axis current at time k; i q (k) is the q-axis current at time k.

4. The six-phase permanent magnet synchronous motor parameter-less model predictive current control system according to claim 3, characterized in that, The mathematical model of the current prediction module is as follows: In the formula, V x x = 1~12 are 12 virtual control voltage vectors; they are composed of two voltage sub-vectors, and the two voltage sub-vectors correspond to V x_L1 and V x_L2 respectively; V x_L1 and V x_L2 are the same in the αβ subspace direction and opposite in the xy subspace direction; the functional expression of V x is as follows: V x = 0.268*V x_L2 + 0.732*V x_L1 ; In the formula, is the d-axis current prediction value for the k+1 instant; is the q-axis current prediction value for the k+1 instant; is the d-axis current prediction value for the k+2 instant; is the q-axis current prediction value for the k+2 instant; V dx is a virtual control voltage vector V x a voltage component in the d-axis; V qx is a virtual control voltage vector V x a voltage component on the q-axis.

5. The six-phase permanent magnet synchronous motor parameter-less model predictive current control system according to claim 4, characterized in that, The mathematical model of the cost function module is as follows: In the formula, g is a cost function; i d * is the d-axis current reference value; i q * is the q-axis current reference value; is the d-axis current prediction value for the k+2 instant; is the q-axis current prediction value for the k+2 instant.

6. The six-phase permanent magnet synchronous motor parameter-less model predictive current control system according to claim 3, characterized in that, The hyper-local model parameter estimation module comprises a hyper-local model d-axis parameter estimation module and a hyper-local model q-axis parameter estimation module; When V d When (k - 1) is zero, the mathematical model of the hyperlocal model d-axis parameter estimation module is: When V d (k-1) is not zero and V d (k-1) is not equal to V d (k-2); the mathematical model of the hyperlocal model d-axis parameter estimation module is: When V d (k-1) is not zero and V d (k-1) is equal to V d (k-2), the mathematical model of the hyperlocal model d-axis parameter estimation module is: When V q When (k - 1) is zero, the mathematical model of the hyperlocal model q-axis parameter estimation module is: When V q (k-1) is not zero and V q (k-1) is not equal to V q (k-2), the mathematical model of the hyperlocal model q-axis parameter estimation module is: When V q (k-1) is not zero and V q (k-1) is equal to V q (k-2), the mathematical model of the hyperlocal model q-axis parameter estimation module is: In the formula, α is a first-order low-pass filter coefficient; F 0_d (k-1) is the disturbance parameter of the d-axis corresponding to the k-1 control period super-local model; F 0_q (k-1) is the disturbance parameter of the q-axis corresponding to the k-1 control period super-local model; A d (k-1) is a gain parameter of the d-axis corresponding to the k-1 control period hyperlocal model; A q (k-1) is a gain parameter of the q-axis corresponding to the k-1 control period hyperlocal model; Δi d (k) is the d-axis current difference at time k; Δi q (k) is the q-axis current difference at time k; Δi d (k-1) is the d-axis current difference at time k-1; Δi q (k-1) is the q-axis current difference at time k-1; V d (k-1) is the k-1 control period d-axis control voltage; V q (k-1) is a k-1 control period q-axis control voltage; V d (k-2) is a k-2 control period q-axis control voltage; V q (k-2) is a k-2 control period q-axis control voltage.

7. The six-phase permanent magnet synchronous motor parameter-less model predictive current control system according to claim 1, wherein, The speed controller comprises a speed PI controller, and the mathematical model of the speed PI controller is as follows: wherein: e n = n * - n; In the formula, i q * is the q-axis current reference value; k nP is a proportional gain of the rotation speed PI controller; k nI is an integral gain of the rotation speed PI controller; s is a complex variable; e n is the difference between the given rotational speed and the actual rotational speed; n is an actual speed; n * For a given rotational speed.

8. A parameter-free model predictive current control method for a six-phase permanent magnet synchronous motor, the method comprising a PWM generation module and an inverter, the PWM generation module being configured to output PWM signals to control the operation of the inverter, and the inverter being configured to drive the motor; characterized in that, The method further comprises a sampling module, a coordinate transformation module, a current difference calculation module, a hyper-local model, a hyper-local model parameter estimation module, a speed controller, and a cost function module; the hyper-local model comprises a time delay compensation module and a current prediction module; The sampling module is configured to sample stator phase currents of each control period; the sampling module outputs stator phase current signals to the coordinate transformation module; The coordinate transformation module is configured to convert the stator phase currents into d-axis and q-axis currents in a dq coordinate system; the coordinate transformation module outputs the d-axis and q-axis currents to the current difference calculation module, the hyper-local model parameter estimation module, and the time delay compensation module, respectively; The current difference calculation module is configured to calculate d-axis and q-axis current differences between adjacent time instants; the current difference calculation module outputs calculation results to the hyper-local model parameter estimation module; The super local model parameter estimation module is configured to estimate parameters of the super local model according to the known current difference of the d and q axes and the vector control voltage, and to output the parameters of the super local model to the delay compensation module and the current prediction module respectively; The delay compensation module is configured to predict the current value of the motor at the k+1 moment under the action of the control voltage of the motor at the k control period; The delay compensation module is configured to output the current prediction value at the k+1 moment to the current prediction module; The current prediction module is configured to predict the d and q axis current values at the k+2 moment under the action of the 12 virtual control voltage vectors respectively, and to output the d and q axis current prediction values at the k+2 moment corresponding to the 12 virtual control voltage vectors to the cost function module; The speed controller is configured to obtain the current reference value through the given speed; The speed controller is configured to input the given speed value and output the current reference value to the cost function module; The cost function module is used to select the optimal vector control voltage according to the error between the current reference value and the current prediction value corresponding to each virtual control voltage vector; the cost function module inputs the current reference value, and respectively calculates the error between the current prediction value at the k+2 moment corresponding to each virtual control voltage vector and the current reference value, and screens the virtual control voltage vector with the minimum cost function value from the 12 virtual control voltage vectors as the k+1 control period control voltage vector V opt , which is sent to the PWM generation module, so that the cost function module converts V opt into the d-axis component and the q-axis component and then sends them to the hyperlocal model parameter estimation module, the delay compensation module and the current prediction module respectively. opt The PWM generation module is used to generate the PWM signal corresponding to the control voltage vector V 9. The six-phase permanent magnet synchronous motor parameter-less model predictive current control method according to claim 8, characterized in that, The method comprises the following steps: Step 1: constructing the mathematical models of the coordinate transformation module, the current difference calculation module, the delay compensation module, the current prediction module, the super local model parameter estimation module, the speed controller, and the cost function module; setting the current control period as the k control period, setting the historical data of the control periods before the k control period, including the historical data of the d and q axis currents of the motor and the d and q axis control voltages of the motor, and assigning initial values to the model parameters of the delay compensation module and the current prediction module; Step 2: obtaining the phase currents of the motor, the rotor position electrical angle, and the motor speed through the sampling module, and obtaining the d and q axis actual currents at each moment through the coordinate transformation method; Step 3: calculating the d axis current difference and the q axis current difference between the k moment and the k-1 moment, and the d axis current difference and the q axis current difference between the k-1 moment and the k-2 moment using the mathematical model of the current difference calculation module; Step 4: obtaining the current prediction value at the k+1 moment from the d and q axis actual currents at the k moment and the d and q axis control voltages at the k control period using the mathematical model of the delay compensation module; In the formula: is the d-axis current prediction value for the k+1 instant; is the q-axis current prediction value for the k+1 instant; F 0_d (k) is a disturbance parameter of the d-axis corresponding to the k control period super-local model; F 0_q (k) is the disturbance parameter of the q-axis corresponding to the k control period super-local model; A d (k) is a gain parameter of the d-axis corresponding to the k control period hyperlocal model; A q (k) is a gain parameter of the q-axis corresponding to the k control period hyperlocal model; V d (k) is a k control period d-axis control voltage; V q (k) is a k control period q-axis control voltage; i d (k) is the d-axis current at time k; i q (k) is the q-axis current at time k; Step 5: Let V x These correspond to 12 virtual control voltage vectors, x = 1 to 12; V dx For V x The voltage component on the d-axis, V qx For virtual control voltage vector V x Voltage component on the q-axis; V x It is synthesized from two voltage vector components, which correspond to V respectively. x_L1 and V x_L2 V x_L1 and V x_L2 The directions are the same in the αβ subspace and opposite in the xy subspace; V x The function expression is as follows: V x = 0.268*V x_L2 + 0.732*V x_L1 ; The mathematical model of the current prediction module is as follows: the current prediction value at the k+1 moment and the virtual control voltage vector V x The voltage component on the q-axis is calculated to obtain the current prediction value at the k+2 moment. In the formula: is the d-axis current prediction value for the k+1 instant; is the q-axis current prediction value for the k+1 instant; is the d-axis current prediction value for the k+2 instant; is the q-axis current prediction value for the k+2 instant; V dx is a virtual control voltage vector V x a voltage component in the d-axis; V qx is a virtual control voltage vector V x a voltage component on the q-axis; Step 6: estimating the parameters of the super local model according to the known current difference and the corresponding vector control voltage, and updating the model parameters of the delay compensation module and the current prediction module using the mathematical model of the local model parameter estimation module; Step 7: The speed controller sets a speed PI controller, sets the d-axis current reference value to 0, sets the parameters of the speed PI controller, and obtains the q-axis current reference value i q * ; Step 8: The cost function module inputs the q-axis current reference value i from the speed controller q * and the current prediction value at the k+2 time instant from the current prediction module, filters the optimal control voltage vector V opt (k+1) for the k+1 control period, sends V opt (k+1) to the PWM generation module; converts V opt (k+1) into the corresponding d-axis component V d (k+1) and q-axis component V q (k+1) to the hyperlocal model parameter estimation module, the delay compensation module and the current prediction module; Step 9: increasing k by 1, updating the values of the d and q axis current differences and the d and q axis control voltages, and returning to step 2.

10. The six-phase permanent magnet synchronous motor parameter-free model predictive current control method according to claim 9, characterized in that, In step 1, the super local model parameter estimation module is set as a super local model d axis parameter estimation module and a super local model q axis parameter estimation module, and the mathematical models of the super local model d axis parameter estimation module and the super local model q axis parameter estimation module are constructed as follows: When V d When (k - 1) is zero, the mathematical model of the hyperlocal model d-axis parameter estimation module is: When V d (k-1) is not zero and V d (k-1) is not equal to V d (k-2); the mathematical model of the hyperlocal model d-axis parameter estimation module is: When V d (k-1) is not zero and V d (k-1) is equal to V d (k-2), the mathematical model of the hyperlocal model d-axis parameter estimation module is: When V q When (k - 1) is zero, the mathematical model of the hyperlocal model q-axis parameter estimation module is: When V q (k-1) is not zero and V q (k-1) is not equal to V q (k-2), the mathematical model of the hyperlocal model q-axis parameter estimation module is: When V q (k-1) is not zero and V q (k-1) is equal to V q (k-2), the mathematical model of the hyperlocal model q-axis parameter estimation module is: In the formula: α is a first-order low-pass filter coefficient; F 0_d (k-1) is the disturbance parameter of the d-axis corresponding to the k-1 control period super-local model; F 0_q (k-1) is the disturbance parameter of the q-axis corresponding to the k-1 control period super-local model; A d (k-1) is a gain parameter of the d-axis corresponding to the k-1 control period hyperlocal model; A q (k-1) is a gain parameter of the q-axis corresponding to the k-1 control period hyperlocal model; Δi d (k) is the d-axis current difference at time k; Δi q (k) is the q-axis current difference at time k; Δi d (k-1) is the d-axis current difference at time k-1; Δi q (k-1) is the q-axis current difference at time k-1; V d (k-1) is a k-1 control period d-axis control voltage; V q (k-1) is a k-1 control period q-axis control voltage; V d (k-2) is a k-2 control period q-axis control voltage; V q (k-2) is a k-2 control period q-axis control voltage.

Images