Model prediction control system and method for minimizing loss of magnetic flux adjustable motor

Abstract

The invention discloses a model prediction control system and method for minimizing loss of a magnetic flux adjustable motor. The system comprises a magnetic flux adjustable motor module, a three-phase inverter module, a coordinate change module, an active current calculation module, a current prediction module, a value function minimization module, a positive and negative pulse detection module, a double-bridge arm inverter circuit module and a rotating speed loop regulator module. The method comprises the following steps: constructing a magnetic flux adjustable motor mathematical model considering iron loss; obtaining a current prediction model considering iron loss based on the magnetic flux adjustable motor loss model; from the perspective of d-axis and q-axis active current, constructing a current constraint relation under the condition of loss minimization; and constructing a value function. According to the invention, the efficiency optimization of the system is improved; the problems of large calculation burden, increased solving error and the like caused by high-order equation solving are avoided; the efficiency optimization of the whole motor driving system is realized; the method is high in response speed, strong in anti-interference capability, small in torque ripple and strong in current harmonic suppression.

Description

Technical Field

[0001] The present invention relates to the field of motor control, and in particular, to a model predictive control system and method for minimizing the losses of a flux-adjustable motor. Background Art

[0002] Permanent magnet synchronous motors have the advantages of a small speed regulation extreme ratio, a strong overload capacity, a fast dynamic response speed, etc. With the improvement of the performance of rare earth permanent magnet materials, the cost of permanent magnet synchronous motors has gradually decreased, and with the improvement of the performance of permanent magnet synchronous motor AC drive systems such as digital signal processors, permanent magnet synchronous motor control technology plays an extremely important role in real life. However, the speed regulation range of traditional permanent magnet synchronous motors is limited by the inherent characteristics of the motor flux linkage, inductance, etc. Moreover, during power generation operation, it is difficult to extinguish the magnetic field during a fault, which limits its application range. Therefore, a flux-adjustable motor is adopted, and a permanent magnet material with high remanence and low coercivity, such as aluminum nickel cobalt (Al Ni Co), can be changed from a non-magnetic state to a magnetic state by applying an instantaneous pulse and can maintain the magnitude and direction of the magnetic flux, thereby realizing the control of the motor.

[0003] Flux-adjustable motors are widely used in electric vehicle drives due to their high power density, excellent torque performance, and suitability for high-speed operation. In order to maximize the driving range of electric vehicles, the core energy-consuming device - the motor drive system - should have higher efficiency. Loss minimization control can comprehensively reduce the copper loss and iron loss during motor operation and further improve the motor operation efficiency. Therefore, the loss minimization control of flux-adjustable motors has become a research hotspot.

[0004] At present, the loss minimization control methods for flux-adjustable motors are divided into two categories: loss minimization control based on a search algorithm and loss minimization control based on a loss model. Among them, the loss minimization control method based on a search algorithm obtains the optimal current operating point by adjusting the d-axis current and monitoring the change of the total motor loss in real time. This method does not depend on the motor parameters, but faces problems such as a long search time, poor stability, and easy system oscillation. The loss minimization control method based on a model directly calculates the optimal current reference values of the d and q axes according to the total motor loss model. This method has the advantage of a fast convergence speed and is suitable for occasions where the motor operating conditions change frequently. It can be seen that the existing loss minimization control methods have reduced the losses of flux-adjustable motors to a certain extent, but need to obtain the optimal reference current by solving a fourth-order nonlinear equation by means of analytical methods, numerical methods, or approximation methods, and these methods face problems such as a large amount of calculation and a large approximation error. Summary of the Invention

[0005] Object of the Invention: The object of the present invention is to provide a model predictive control system and method for minimizing the losses of a flux-adjustable motor.

[0006] Technical solution: The model predictive control system for minimizing the losses of a flux-adjustable motor according to the present invention includes a flux-adjustable motor module, a three-phase inverter module, a coordinate transformation module, an active current calculation module, a current prediction module, a cost function minimization module, a positive and negative pulse detection module, a two-bridge arm inverter circuit module, and a speed loop regulator module. The stator currents i a 、i b 、i c of the flux-adjustable motor are converted into current values i d and i q , and are respectively input into the current prediction module and the active current calculation module. The positive and negative pulse detection module outputs the number of positive and negative pulses n1 and n2 to the active current calculation module. The active current calculation module outputs the active current values i d 、i q based on the positive and negative pulses n1, n2, and the current values i wd (k) and i wq (k), and inputs them into the current prediction module; The current prediction module predicts the predicted active current values i d 、i q and the active current values i wd (k) and i wq (k), and predicts the predicted active current values i wd (k + 1) and i wq (k + 1) at the next moment, and inputs the current prediction values i wd (k + 1) and i wq (k + 1) into the cost function minimization module. The cost minimization module selects the switching state corresponding to the minimum cost function in the three-phase inverter module based on i wq output by the speed loop regulator module and the predicted active current values i wd (k + 1) and i wq (k + 1) at the next moment to control the flux-adjustable motor module.

[0007] Further, the stator currents i a 、i b 、i c of the flux-adjustable motor are converted into current values i d and i q through the coordinate transformation module.

[0008] The implementation of the model predictive control system for minimizing the losses of the flux-adjustable motor according to the present invention is characterized by including the following steps:

[0009] (1) Construct a mathematical model of the flux-adjustable motor considering iron loss;

[0010] (2) Obtain a current prediction model considering iron loss based on the loss model of the flux-adjustable motor;

[0011] (3) Starting from the perspective of the active current on the d and q axes, construct the current constraint relationship under the condition of minimizing losses;

[0012] (4) Construct the value function.

[0013] Furthermore, the step (1) includes:

[0014] According to the relationship between magnetic flux, inductance, and magnetic linkage:

[0015]

[0016] ψ = Nφ

[0017] Where: L is the inductance, ψ is the magnetic linkage, is the magnetic flux, i is the current, and N is the number of turns;

[0018] When adjusting the magnetic flux through the magnetic modulation pulse, the values of the corresponding d and q axis inductances and the permanent magnet magnetic linkage will also change. Therefore, L and ψ f After being adjusted by the positive and negative pulses n1 and n2, their expressions are:

[0019] L = L0 + n1ΔL - n2ΔL

[0020] ψ f = ψ0 + n1Δψ - n2Δψ

[0021] Where, L0 is the initial parameter of the inductance, ΔL is the change amount of the inductance parameter corresponding to one magnetic modulation pulse, ψ0 is the initial parameter of the magnetic linkage, Δψ is the change amount of the magnetic linkage parameter corresponding to one magnetic modulation pulse, and n1 and n2 are the positive and negative pulse numbers.

[0022] Furthermore, the voltage equation of the flux - adjustable motor mathematical model considering iron loss in the step (1) is:

[0023]

[0024] The stator current equation during the steady - state operation of the motor is:

[0025]

[0026] The expression of the electromagnetic torque Te is:

[0027]

[0028] Where, u d and u q are the stator d and q axis voltages respectively, i d and i q are the stator d and q axis voltages and currents respectively, i cd and icq are the iron loss current components of the d and q axes, respectively, and i wd and i wq are the active current components participating in the electromagnetic torque synthesis, and R s is the stator resistance, and L d and L q are the d and q axis inductances, respectively, and ψ f is the rotor permanent magnet flux linkage, W is the mechanical angular velocity, and n p is the number of pole pairs, and Te is the electromagnetic torque value.

[0029] Furthermore, the step (2) includes:

[0030] Obtain a current prediction model considering iron loss based on the flux adjustable motor loss model, and the expression is:

[0031]

[0032] where the active current i wdq (k) cannot be directly measured, so it needs to be indirectly calculated through the stator current i dq ; L d is the d axis inductance, i wd (k) and i wq (k) are the active current values at the current moment; T s is the sampling time; L d and L q are the d and q axis inductance values, respectively; R s is the stator resistance value.

[0033] Furthermore, the stator current i dq can be calculated through the stator current equation:

[0034]

[0035] where:

[0036]

[0037] The active current at the current moment is:

[0038]

[0039] where i wd (k + 1) and i wq (k + 1) are the predicted d and q axis active current values at the next moment by the current prediction model; i d and i q are the stator d and q axis voltages and currents; u d and u q are the stator d and q axis voltages; ω is the motor speed value; ψ fis the rotor permanent magnet flux linkage; a and b are self-defined coefficients.

[0040] Further, the losses of the motor in step (3) include stator winding losses, core losses, mechanical losses, and stray losses. The expression for the total losses of the motor is:

[0041]

[0042] where: P loss is the total loss value; P cu is the copper loss value; P Fe is the iron loss value.

[0043] Further, the construction relationship of the current constraint in step (3) is:

[0044] Construct the Lagrangian function and calculate the constraint relationship satisfied between the currents i wq and i wd where λ is the Lagrange multiplier; Te is the torque value; n

[0045]

[0046] is the number of pole pairs; Take the partial derivatives of i p and i wd in the Lagrangian function constructed above with respect to i wq respectively to obtain:

[0047]

[0048] Eliminate λ in the above partial derivatives to obtain the constraint relationship satisfied between the d-axis and q-axis active currents i wq and i wd under the condition of minimum loss:

[0049]

[0050] where the coefficients A, B, C, and D are respectively:

[0051]

[0052] where A, B, and C are all self-defined coefficients.

[0053] Further, the current constraint in step (4) is expressed as:

[0054]

[0055] where i * wq is the q-axis active current given by the speed loop PI regulator, and I maxis the stator current limit value; the first and second terms in the value function enable the motor to meet the minimum loss constraint condition of the motor under a given load; the third term is a penalty factor for limiting the maximum current of the motor. When the predicted current of the motor exceeds the stator current limit value I max , the value function value will tend to infinity.

[0056] Beneficial effects: Compared with the prior art, the present invention has the following remarkable advantages: When the flux-adjustable motor of the present invention operates, the factors that cause changes in values such as inductance and magnetic flux during the flux adjustment process are considered in the calculation of losses, and the minimum loss is jointly obtained, enabling the flux-adjustable motor to operate in an environment of minimum loss, thereby improving the efficiency optimization of the system; there is no need to solve high-order equations to obtain the optimal current value, avoiding problems such as large computational burden and increased solution error caused by solving high-order equations; the efficiency optimization of the entire motor drive system is achieved; fast response speed, strong anti-interference ability, small torque ripple, and strong current harmonic suppression. Description of the Drawings

[0057] Figure 1 is the structural block diagram of the model predictive control system for minimizing the loss of the flux-adjustable motor of the present invention;

[0058] Figure 2 is the structural block diagram of the active current calculation module;

[0059] Figure 3 is the structural block diagram of the current prediction module. Detailed Embodiment

[0060] The technical solution of the present invention will be further described below with reference to the drawings.

[0061] As Figure 1 shown, it is the structural block diagram of the model predictive control system for minimizing the loss of the flux-adjustable motor of the present invention, including a flux-adjustable motor module 1, a three-phase inverter module 2, a coordinate transformation module 3, an active current calculation module 4, a current prediction module 5, a value function minimization module 6, a positive and negative pulse detection module 7, a double-bridge arm inverter circuit module 8, and a speed loop regulator module 9. The stator currents i a , i b , i c of the flux-adjustable motor 1 are converted into current values i d and i q , and are respectively input into the current prediction module 5 and the active current calculation module 4. The positive and negative pulse detection module 7 outputs positive and negative pulse numbers n1 and n2 to the active current calculation module 4. The active current calculation module 4 outputs the active current values i d , i q , and outputs the active current value i wd (k) and i wq(k), and input it into the current prediction module 5; the current prediction module 5 predicts the active current prediction values i d 、i q and the active current value i wd (k) and i wq (k), and predicts the active current prediction values i wd (k + 1) and i wq (k + 1), and inputs the current prediction values i wd (k + 1) and i wq (k + 1) into the value function minimization module 6. The value minimization module 6 selects the switching state corresponding to the minimum value of the value function in the three-phase inverter module 2 according to i * wq output by the speed loop adjustment module 9 and the active current prediction values i wd (k + 1) and i wq (k + 1) to control the flux adjustable motor module 1.

[0062] The model predictive control method for minimizing the loss of the flux adjustable motor according to the present invention specifically includes the following steps:

[0063] (1) Construct a mathematical model of the flux adjustable motor considering iron loss;

[0064] Construct a mathematical model of the flux adjustable motor module considering iron loss. In actual operation, the magnetic flux of the flux adjustable motor will change. According to the relationship between magnetic flux, inductance, and magnetic chain:

[0065]

[0066] ψ = Nφ

[0067] It can be seen that when the magnetic flux changes, the values of the d-axis and q-axis inductances and the permanent magnet magnetic chain will also change. Therefore, after L and ψ f are adjusted by the positive pulse n1 and the negative pulse n2, their expressions are:

[0068] L = L0 + n1ΔL - n2ΔL

[0069] ψ f = ψ0 + n1Δψ - n2Δψ

[0070] The voltage equation of the mathematical model of the flux adjustable motor considering iron loss is:

[0071]

[0072] The stator current equation during the steady-state operation of the motor is:

[0073]

[0074] The expression of the electromagnetic torque Te is as follows:

[0075]

[0076] First, construct the mathematical model of the flux-adjustable motor module considering iron loss; then collect the stator current commands i a 、i b 、i c of the flux-adjustable motor module 1. After passing through the coordinate transformation module 3, the currents id and iq are output; the current values i d and i q are respectively input into the coordinate transformation module 3 and the active current calculation module 4; the positive and negative pulse detection module 7 outputs the positive and negative pulse numbers n1 and n2 to the active current calculation module 4; the active current calculation module outputs the active current values i d 、i q under the input of the positive and negative pulses n1, n2, and the current values i wd (k) and i wq (k), and inputs them into the current prediction module 5; the current prediction module 5 predicts the predicted active current values i d 、i q and the active current values i wd (k) and i wq (k), and predicts the predicted active current values i wd (k + 1) and i wq (k + 1) at the next moment, and inputs the current prediction values i wd (k + 1) and i wq (k + 1) into the value function minimization module 6. The value minimization module 6 selects the switching state corresponding to the minimum value of the value function in the three-phase inverter module 2 according to the i * wq output by the speed loop adjustment module 9 and the predicted active current values i wd (k + 1) and i wq (k + 1) at the next moment to control the flux-adjustable motor module 1.

[0077] (2) Obtain the current prediction model considering iron loss based on the flux-adjustable motor loss model;

[0078] As Figure 2 shown is the structural block diagram of the active current calculation module proposed by the present invention. The active current calculation module 4 calculates the active current values i d (k), i q (k) at the current moment by inputting the stator current values i wd (k), i wq (k) and the positive and negative pulse numbers n1, n2, etc.;

[0079] The differential equations in the d and q coordinate systems are as follows:

[0080]

[0081] Since the active current i wdq (k) cannot be directly measured, it is necessary to indirectly calculate it through the stator current i dq .

[0082] Combining with the stator current equation during the steady-state operation of the motor, we can obtain:

[0083]

[0084] Where:

[0085]

[0086] Therefore, the active current at the current moment output by the active current calculation module 4 is:

[0087]

[0088] As Figure 3 shown is the structural block diagram of the current prediction module proposed by the present invention;

[0089] Based on the active current values i wd (k), i wq (k) output by the active current module 4 and the stator current i d , i q output by the coordinate transformation module 3, the current prediction module 5 predicts the active current prediction values i wd (k + 1), i wq (k + 1) at the next moment.

[0090] The expression of the current prediction module 5 is:

[0091]

[0092] (3) Starting from the perspective of the active current in the d and q axes, construct the current constraint relationship under the condition of minimizing losses;

[0093] The losses of the motor mainly include stator winding losses (copper losses), core losses (iron losses), mechanical losses, and stray losses. Generally, the mechanical losses and stray losses account for a relatively small proportion and are difficult to measure and control. Therefore, the efficiency optimization control generally targets the controllable losses such as copper losses and iron losses. Comprehensively reducing the total losses (the sum of copper losses and iron losses) during the operation of the motor is the optimization goal of the minimum loss control. In order to construct the value function minimization module 6 that meets the condition of minimizing losses, the present invention starts from the perspective of the active current and derives the constraint relationship between the active currents in the d and q axes in the state of minimum losses.

[0094] According to the mathematical model, the expression for the total motor loss is:

[0095]

[0096] In the formula: P loss is the total loss value; P cu is the copper loss value; P Fe is the iron loss value.

[0097] To minimize P loss , the Lagrangian function is constructed as follows:

[0098]

[0099] where λ is the Lagrange multiplier; Te is the torque value; n p is the number of pole pairs.

[0100] Taking the partial derivatives of i wd and i wq in the Lagrangian function constructed above respectively, we get:

[0101]

[0102] Eliminating λ in the above partial derivatives, we can obtain the constraint relation satisfied between the active currents iwq and iwd on the d and q axes under the condition of minimum loss:

[0103]

[0104] where the coefficients A, B, C, and D are respectively:

[0105]

[0106] (4) Construct the value function;

[0107] In order to reduce the motor loss simultaneously without solving high-order equations to obtain the optimal current value, the value function minimization module 6 is constructed by using the motor loss minimization constraint condition as follows:

[0108]

[0109] The first and second terms in the value function make the motor satisfy the motor loss minimization constraint condition under a given load; the third term is a penalty factor used to limit the maximum current of the motor. When the predicted current of the motor exceeds the stator current limit value I max , the value of the value function will tend to infinity; the optimal voltage vector corresponding switch state can be directly selected from the basic voltage vectors by using the value function, thus avoiding the heavy calculations brought by solving the optimal reference current.

[0110] The PI regulator of the rotational speed control loop in the present invention is used to obtain the reference value of the active current on the q-axis. In the current control loop, the active current i dq (k) at the current moment is calculated through the i collected at the current moment wdq (k), and then the active current i under the action of 8 basic voltage vectors at the k+1 moment is calculated by using the prediction model wdq (k+1), and further the switching state corresponding to the voltage vector that minimizes the cost function is calculated to control the flux-adjustable motor drive system.

Claims

1. A model predictive control system for minimizing losses in a flux-adjustable motor, characterized in that: The invention comprises a flux adjustable motor module (1), a three-phase inverter module (2), a coordinate change module (3), an active current calculation module (4), a current prediction module (5), a value function minimization module (6), a positive and negative pulse detection module (7), a dual-bridge arm inverter circuit module (8) and a speed loop regulator module (9). The stator current i of the flux adjustable motor (1) is a 、i b 、i c Convert to current value i d and i q , respectively input into the current prediction module (5) and the active current calculation module (4), the positive and negative pulse detection module (7) outputs the positive and negative pulse numbers n1 and n2 to the active current calculation module (4), and the active current calculation module (4) calculates the current value i according to the positive and negative pulses n1 and n2. d 、i q , output active current value i wd (k) and i wq (k), and input it into the current prediction module (5); the current prediction module (5) calculates the current value i according to the current value i d 、i q and active current value i wd (k) and i wq (k), predict the active current prediction value i at the next moment wd (k+1) and i wq (k+1), and the current prediction value i wd (k+1) and i wq (k+1) is input to the value function minimization module (6), and the value minimization module (6) is based on the i output by the speed loop adjustment module (9). * wq and the predicted active current value i at the next moment wd (k+1) and i wq (k+1), selecting the switch state corresponding to the minimum value function in the three-phase inverter module (2) to control the flux-adjustable motor module (1).

2. The model predictive control system for minimizing losses of a flux-adjustable motor according to claim 1, characterized in that: The stator current i of the flux adjustable motor (1) a 、i b 、i c Converted into current value i through coordinate change module d and i q .

3. A model predictive control method for minimizing the loss of a flux-adjustable motor, which is implemented by the model predictive control system for minimizing the loss of a flux-adjustable motor according to claim 1, characterized in that: The steps include: (1) Construct a mathematical model of a flux-adjustable motor taking into account iron loss; (2) Based on the flux adjustable motor loss model, a current prediction model considering iron loss is obtained; (3) From the perspective of the active current of the d and q axes, the current constraint relationship under the condition of loss minimization is constructed; (4) Construct a value function.

4. The model predictive control method for minimizing the loss of a flux-adjustable motor according to claim 3, characterized in that: The step (1) comprises: According to the relationship between magnetic flux, inductance and magnetic flux: ψ=Nφ Where: L is the inductance, ψ is the magnetic flux, is the magnetic flux, i is the current, and N is the number of turns; When the magnetic flux is adjusted by the magnetic pulse, the corresponding d-axis and q-axis inductances and the permanent magnet flux linkage values ​​will also change, so L and ψ f After adjustment by positive and negative pulses n1 and n2, the expression is: L=L0+n1ΔL-n2ΔL ψ f =ψ0+n1Δψ-n2Δψ Among them, L0 is the initial inductance parameter, ΔL is the change in inductance parameter corresponding to a magnetic modulation pulse, ψ0 is the initial flux parameter, Δψ is the change in flux parameter corresponding to a magnetic modulation pulse, and n1 and n2 are the positive and negative pulse numbers.

5. The model predictive control method for minimizing losses of a flux-adjustable motor according to claim 3, characterized in that: The voltage equation of the mathematical model of the flux-adjustable motor considering the iron loss in step (1) is: The stator current equation of the motor during steady-state operation is: The expression of electromagnetic torque Te is: Among them, u d and u q Stator d and q axis voltages, i d and i q are the stator d and q axis voltage and current respectively, i cd and i cq are the d-axis and q-axis iron loss current components, i wd and i wq are the active current components participating in the synthesis of electromagnetic torque, R s is the stator resistance, L d and L q are the d-axis and q-axis inductances, ψ f is the rotor permanent magnet flux, W is the mechanical angular velocity, n p is the number of pole pairs, Te is the electromagnetic torque value.

6. The model predictive control method for minimizing the loss of a flux-adjustable motor according to claim 3, characterized in that: The step (2) comprises: Based on the flux adjustable motor loss model, the current prediction model considering iron loss is obtained, and the expression is: Among them, the active current i wdq (k) cannot be measured directly, so it is necessary to measure the stator current i dq To calculate indirectly; L d is the d-axis inductance, i wd (k) and i wq (k) is the active current value at the current moment; T s is the sampling time; L d and L q are the d-axis and q-axis inductance values ​​respectively; R s is the stator resistance value.

7. The model predictive control method for minimizing the loss of a flux-adjustable motor according to claim 6, characterized in that: The stator current i dq The stator current equation can be calculated as follows: in: The active current at the current moment is: Among them, i wd (k+1) and i wq (k+1) is the d-axis and q-axis active current values ​​predicted by the current prediction model at the next moment; i d and i q is the stator d and q axis voltage and current; u d and u q is the stator d and q axis voltage; ω is the motor speed value; ψ f is the rotor permanent magnet flux; a, b, c are self-determined coefficients.

8. The model predictive control method for minimizing losses of a flux-adjustable motor according to claim 3, characterized in that: The loss of the motor in step (3) includes stator winding loss, core loss, mechanical loss and stray loss. The total loss of the motor is expressed as: Where: P loss is the total loss value; P cu is the copper loss value; P Fe is the iron loss value.

9. The model predictive control method for minimizing losses of a flux-adjustable motor according to claim 3, characterized in that: The current constraint relationship in step (3) is constructed as follows: Construct the Lagrangian function and calculate the current i wq and i wd The constraint relationship between them is satisfied. Where, λ is the Lagrange multiplier; Te is the torque value; n p is the pole logarithm; i in the Lagrangian function constructed in the above formula wd with i wq Find the partial derivatives separately and get: Eliminate the partial derivative of the above formula to get λ and get the d and q axis active current i under the condition of loss minimization wq and i wd The constraint relationship satisfied between them is: Among them, the coefficients A, B, C and D are: Among them, A, B, and C are all custom coefficients.

10. The model predictive control method for minimizing losses of a flux-adjustable motor according to claim 3, characterized in that: The current constraint relationship in step (4) is expressed as: Among them, i * wq is the q-axis active current given by the speed loop PI regulator, I max is the stator current limit value; the first and second terms in the cost function make the motor meet the minimum motor loss constraint under a given load; the third term is a penalty factor used to limit the maximum current of the motor. When the predicted current of the motor exceeds the stator current limit value I max , the value function will tend to infinity.

Images