Multi-target intelligent optimization efficient control method for cascaded multi-level electric drive system

Abstract

The invention discloses a multi-target intelligent optimization efficient control method for a cascaded multilevel electric drive system, and the method comprises the steps: constructing a motor broadband domain loss model and a converter loss model, and obtaining an electric drive system loss model after unifying current variables; deriving a motor torque ripple model taking a specific secondary current harmonic component as a component based on a magnetic field modulation principle; establishing an electric drive system loss multi-objective optimization model which takes a fundamental current angle and a specific harmonic current component as two-dimensional optimization variables and takes torque ripple performance into account; and performing particle optimization configuration in the two-dimensional optimization variables of the fundamental current angle and the specific harmonic current component by applying a particle swarm algorithm, solving the multi-objective optimization model, and obtaining the optimal fundamental current angle and the optimal harmonic current component setting. According to the method, the multi-target and multi-dimensional complex problem of the electric drive system is solved by adopting the particle swarm optimization, global optimization can be carried out from the multi-control-dimension angle, and the defect that a traditional optimization algorithm is converged to a local optimal point is overcome.

Description

Technical Field

[0001] The present invention belongs to the field of motor and control theory application, and in particular relates to a multi-objective intelligent optimization and efficient control method for a cascaded multi-level electric drive system. Background Art

[0002] Permanent magnet synchronous motors (PMSMs), with their advantages of high power density, high efficiency, and excellent control performance, have been widely used in electric drives, particularly high-performance electric drives. In high-voltage, high-power applications, multilevel converters better meet motor drive requirements, while cascaded multilevel topologies are widely used due to their simpler structure and easier modular expansion. Therefore, with the increasing demand for high-voltage, high-power applications, research on efficiency optimization of electric drive systems based on wide-band loss models for motors and cascaded multilevel converter loss models is of practical significance. While optimizing motor control for high efficiency, other aspects of the electric drive system's operating performance must also be considered. For example, motor torque ripple significantly impacts motor mechanical stress, directly impacting its reliable operation. Therefore, research on efficient, multi-objective control of electric drive systems is also of practical significance. Furthermore, reducing motor torque ripple requires the active injection of harmonic currents into the system. Therefore, efficiency optimization requires not only d-axis current control but also optimization of harmonic current injection, making it a multi-dimensional control problem. Furthermore, the multi-objective function and multi-dimensional control of motors present the complex characteristics of multi-dimensional and multi-peak optimization control.

[0003] Most existing optimization control strategies that consider the loss model of electric drive systems use traditional search algorithms, such as the golden section algorithm. Most of them approximate extreme points through the interval contraction principle. They are suitable for single-objective optimization problems, but it is difficult to directly handle multiple objectives. The selection of the initial point is very important, and when faced with multiple extreme value problems, it is easy to fall into the local optimal point, and the global search capability is poor. Summary of the Invention

[0004] The purpose of the present invention is to provide a multi-objective intelligent optimization and efficient control method for a cascaded multi-level electric drive system. The particle swarm algorithm is used to solve the multi-objective and multi-dimensional complex problems of the electric drive system. It can not only perform global optimization from the perspective of multiple control dimensions, but also avoid the defect of traditional optimization algorithms converging to local optimal points.

[0005] In order to achieve the above object, the solution of the present invention is:

[0006] A multi-objective intelligent optimization and efficient control method for a cascaded multi-level electric drive system comprises the following steps:

[0007] Step 1: Construct a wide-frequency domain loss model for the motor and a converter loss model, unify the current variables, and obtain the electric drive system loss model;

[0008] Step 2: Based on the optimization of the fundamental current angle (i.e., the angle between the fundamental quadrature and direct axis current reference values), a cascaded multi-level permanent magnet synchronous electric drive loss optimization model is designed.

[0009] Step 3: Establish a permanent magnet synchronous motor torque ripple model based on the magnetic field modulation principle, obtain a mathematical mapping relationship between current harmonic components and torque ripple components, and derive a motor torque ripple model with specific sub-current harmonic components as components;

[0010] Step 4: Based on the motor torque ripple model, design a motor torque ripple suppression method based on injection of specific sub-current harmonic components;

[0011] Step 5: Based on the electric drive system loss model constructed in step 1 and the motor torque ripple model obtained in step 3, a multi-objective optimization model for electric drive system loss is established, which takes the fundamental current angle and specific subharmonic current components as two-dimensional optimization variables and takes into account torque ripple performance.

[0012] Step 6: Based on the multi-objective optimization model of electric drive system loss that takes into account torque ripple performance obtained in Step 5, a particle swarm algorithm is applied to perform particle optimization configuration in the two-dimensional optimization variables of the fundamental current angle and the specific subharmonic current component;

[0013] In step 7, based on the particle optimization configuration in step 6, the multi-objective optimization model of the electric drive system loss established in step 5 is solved to obtain the optimal fundamental current angle and optimal harmonic current component settings of the electric drive system.

[0014] In step 1 above, the expression of the electric drive system loss model is,

[0015] P L =minP Cu +P Fe,f (i od )+P Fe,h (i od )+P inv (i od )

[0016] Among them, P L is the electric drive system loss, P Cu is the copper loss, P Fe,f is the motor fundamental frequency loss, P Fe,h is the high-frequency loss of the motor, P inv is the converter loss.

[0017] In step 1 above, the motor wide-frequency domain loss model is constructed as follows:

[0018]

[0019] Among them, u df and uqf ,u dh and u qh are the dq axis fundamental frequency voltage and dq axis high frequency voltage respectively, R s is the stator resistance; i df and i qf ,i dh and i qh are the dq axis fundamental frequency current and dq axis high frequency current respectively, p is the derivative symbol, L d and L q are the dq axis inductance, ψ p is the permanent magnet flux, ω e is the electrical angular frequency; R i is the equivalent iron loss resistance;

[0020] Among them, the equivalent iron loss resistance R i Including the fundamental frequency iron loss equivalent resistance R c , iron loss equivalent resistance R at 1 times equivalent switching frequency f1 and the equivalent resistance R of the iron loss at 2 times the equivalent switching frequency f2 , the converter loss model is constructed as,

[0021]

[0022] Among them, P inv is the converter loss, V on is the collector-emitter voltage when conducting, R on is the on-resistance, I m is the three-phase current amplitude; e on ,e off ,e rec are the reference values ​​of turn-on loss, turn-off loss and reverse recovery loss, respectively. c is the switching frequency, V dc is the DC side voltage of the H-bridge.

[0023] In the above step 2, the optimization model of the electric drive system loss model is first constructed as follows:

[0024] Optimization target P L =minP Cu +P Fe,f (i od )+P Fe,h (i od )+P inv (i od )

[0025] Equality constraint T e =1.5p n i oq [ψ p +(L d -Lq )i od ]

[0026] Inequality constraints

[0027] Among them, P L is the electric drive system loss, P Cu is the copper loss, P Fe,f is the motor fundamental frequency loss, P Fe,h is the high-frequency loss of the motor, P inv is the converter loss, i od is the magnetizing current; T e is the torque, p n is the pole pair number, ψ p is the permanent magnet flux, L d and L q are the d-axis and q-axis inductances respectively; U sm is the maximum voltage amplitude for safe operation of the motor, ω e is the electrical angular frequency; i s is the stator current, i d ,i q are dq axis current, i sm The maximum current value for safe operation of the motor; i od,min is the minimum value of magnetizing current, i od,max is the maximum value of magnetizing current; f sw is the switching frequency, f sw,min is the minimum switching frequency for normal operation of the motor, f sw,max is the maximum switching frequency for normal operation of the motor; its optimization method is to control the fundamental current angle to obtain the minimum value of the objective function.

[0028] In step 3 above, the torque ripple model of the permanent magnet synchronous motor is obtained based on the magnetic field modulation principle as follows:

[0029]

[0030] Among them, P is the constant term composed of the motor structure and electromagnetic constant, F f is the permanent magnet potential, F a is the armature magnetic potential, θ is the motor space angle, and t is the time; according to the magnetic field modulation theory, the torque generated is divided into three parts: the torque T generated by the rotor magnetic field modulated by the stator slots ff , the torque generated by the interaction between the stator and rotor magnetic fields (permanent magnet torque) T fa , the component T of the armature magnetic field modulated by the stator slots aa , for T aa Its size is independent of the rotor position and it does not actually contribute to torque.

[0031] In step 3 above, the mathematical mapping relationship between the current harmonic components and the torque ripple components is obtained, and the motor torque ripple model with the specific sub-current harmonic components as components is derived as follows:

[0032]

[0033] Among them, T 12 is the total 12 torque pulsation amplitudes, ω is the electrical angle, Generates 12 torque ripple phases for the fundamental wave; K 1_12 is the constant of torque pulsation generated by fundamental current, I1 is the square root sum of the motor dq axis current amplitude, is the fundamental torque phase, It is the phase difference between the fundamental wave and the 12th order torque ripple.

[0034] In step 4 above, the expression for suppressing the 12th torque ripple based on harmonic current injection is:

[0035]

[0036] Among them, T 1_12 、T 11_12 、T 12 They are the fundamental wave generation, 11th current harmonic generation and the total 12th torque ripple amplitude, They are respectively the fundamental wave generation, the 11th current harmonic generation and the total 12th torque ripple phase, and ω is the electrical angle.

[0037] In step 5 above, the penalty function pen is introduced to construct the multi-objective optimization model of electric drive system loss, and the optimization target Fitness = minP Cu +P Fe,f (i od )+P Fe,h (i od )+P inj_loss (i inj )+pen(I(C1,T 12 ))

[0038] Equality constraint T e =1.5p n i oq [ψ p +(L d -L q )i od ]

[0039] Inequality constraints

[0040] Among them, Fitness is the multi-objective optimization objective function, P Cu is the motor copper loss PCu , P Fe_f is the fundamental frequency iron loss, P Fe_h is the high frequency iron loss, P inj_loss is the loss caused by the injected current harmonics, pen is the penalty function, C1 is a constant; I is the discriminant function, if the 12th order torque ripple T 12 If the torque pulsation amplitude is greater than the torque pulsation boundary amplitude C1, the I output is 1, otherwise it is 0.

[0041] In step 6 above, the particle optimization configuration is performed in the two-dimensional optimization variables of the fundamental current angle and the specific subharmonic current component as follows:

[0042] v n (i′ d )=wv n (i d )+c1r1[p n (i d )-x n (i d )]+c2r2[g best -x n (i d )]

[0043] x n (i' d )=x n (i d )+v n (i′ d )

[0044] v n (i' 11_mag )=wv n (i 11_mag )+c1r1[p n (i 11_mag )-x n (i 11_mag )]+c2r2[g best -x n (i 11_mag )]

[0045] x n (i' 11_mag )=x n (i 11_mag )+v n (i 11_mag )

[0046] v n (i' 11_phase )=wv n (i 11_phase )+c1r1[p n (i 11_phase )-xn (i 11_phase )]+c2r2[g best -x n (i 11_phase )]

[0047] x n (i' 11_phase )=x n (i 11_phase )+v n (i 11_phase )

[0048] Among them, v n (i′ d ) is the updated velocity of the particle in the d-axis current dimension, p n (i d ) is the optimal position of the nth particle under the d-axis current dimension, g best is the optimal position of the group under the d-axis current dimension, v n (i′ 11_mag ) is the updated velocity of the particle in the 11th current harmonic amplitude dimension, p n (i 11_mag ) is the optimal position of the nth particle in the 11th harmonic amplitude dimension, x n (i′ 11_mag ) is the updated position of the particle in the 11th current harmonic amplitude dimension, v n (i′ 11_phase ) is the updated velocity of the particle in the 11th current harmonic angle dimension, p n (i 11_phase ) is the optimal position of the nth particle in the 11th harmonic angle dimension, x n (i′ 11_phase ) is the updated position of the particle in the 11th current harmonic angle dimension; v n (i d ),x n (i d ) are the velocity and position of the nth particle in the d-axis current dimension; v n (i 11_mag ),x n (i 11_mag ) are the velocity and position of the nth particle under the dimension of the amplitude of the injected 11th current harmonic; v n (i 11_phase ),x n (i 11_phase ) are the velocity and position of the nth particle under the phase dimension of the injected 11th current harmonic; w is the inertia weight, c1 and c2 are learning factors; r1 and r2 are uniform random numbers in the range of [0, 1].

[0049] In step 7 above, multiple variables of the multi-objective optimization model are updated and positioned in the multi-dimensional space, thereby searching for the global optimal point, which is the optimal operating current and the optimal injection current of the electric drive system.

[0050] After adopting the above scheme, the present invention constructs an integrated loss model of the electric drive system based on the wide-frequency domain loss model of the permanent magnet synchronous motor and the loss model of the cascaded multi-level converter, and at the same time constructs a mathematical model of the torque pulsation of the permanent magnet synchronous motor. Based on the torque pulsation model, an optimization strategy for suppressing torque pulsation by harmonic current injection is proposed, and a penalty function is introduced to construct multiple optimization objectives of torque pulsation and system loss. A multi-objective optimization control method of the electric drive system based on the particle swarm intelligent optimization algorithm is proposed, and the torque pulsation and overall loss of the electric drive system are optimized in multiple control dimensions according to the two dimensions of d-axis current and harmonic current.

[0051] Compared with the prior art, the present invention has the following advantages:

[0052] (1) The present invention establishes a loss model for the electric drive system and uses linear superposition to perform wide-frequency domain modeling of the motor loss under converter power supply. Compared with the traditional fundamental frequency iron loss permanent magnet synchronous motor equivalent circuit, this model takes into account the high-frequency loss of the motor and is more accurate.

[0053] (2) The particle swarm two-dimensional optimization control strategy based on the wide-band loss model of the cascaded multi-level electric drive system proposed in this invention has a greatly enhanced global search capability compared to the traditional optimization search algorithm. When faced with multi-dimensional and multi-peak complex problems, it can converge to the global optimal point more accurately.

[0054] (3) Based on the multi-objective optimization model of the electric drive system, the present invention applies the particle swarm algorithm for optimization. By setting learning parameters and adjusting the algorithm optimization strategy, the adaptive weights can ensure that the global search capability is strong and the convergence is stable. By introducing the penalty function, a particle swarm multi-objective optimization algorithm is constructed. The search algorithm is applied through a unified objective function format, which solves the problems of the traditional search algorithm well and avoids the optimization effects brought by complex multi-objective algorithms. It has a wider application significance. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 This is the equivalent circuit diagram of the permanent magnet synchronous motor without considering the eddy current reaction, that is, without considering the motor iron loss;

[0056] Where, (a) is the d-axis, (b) is the q-axis;

[0057] Among them, 1.1 is the stator resistance, 1.2 is the coupling potential of the q-axis to the d-axis, 1.3 is the d-axis stator voltage, 1.4 is the d-axis inductance, 1.5 is the coupling potential of the d-axis to the q-axis, 1.6 is the q-axis inductance, and 1.7 is the q-axis stator voltage;

[0058] Figure 2 This is the equivalent circuit diagram of a permanent magnet synchronous motor that takes into account the wide-frequency losses of the motor caused by the high-frequency power supply effect of the converter;

[0059] Where, (a) is the d-axis, (b) is the q-axis;

[0060] Where, 2.1 is the stator winding resistance, 2.2 is the coupling potential of the q-axis to the d-axis, 2.3 is the voltage applied to the stator d-axis when powered by the inverter, 2.4 is the equivalent resistance of the motor's wide-band loss, 2.5 is the d-axis inductance, 2.6 is the coupling potential of the d-axis to the q-axis, 2.7 is the voltage applied to the stator q-axis when powered by the inverter, and 2.8 is the q-axis inductance.

[0061] Figure 3 It is a flow chart of a multi-objective intelligent optimization and efficient control method for a cascaded multi-level electric drive system;

[0062] Among them, 3.1 is the step of establishing the electric drive system loss model, 3.2 is the step of proposing the electric drive loss optimization control method, 3.3 is the step of proposing the permanent magnet synchronous motor torque pulsation model, 3.4 is the step of proposing the torque pulsation suppression strategy based on harmonic injection, 3.5 is the step of constructing the electric drive loss model taking into account the torque pulsation, 3.6 is the step of the particle swarm optimization algorithm, and 3.7 is the multi-objective and multi-dimensional optimization algorithm of the electric drive system;

[0063] Figure 4 It is the flow chart of particle swarm optimization algorithm;

[0064] Among them, 4.1 is to initialize the position and speed of the particle swarm, 4.2 is to calculate the objective function value of each particle, 4.3 is to calculate the minimum objective function value of each individual particle, 4.4 is to calculate the minimum objective function value in the swarm, 4.5 is to optimize the speed and position of the particles, 4.6 is to execute the boundary conditions, 4.7 is to determine whether the convergence conditions are met, if they are met, 4.8 is to output the results, if not, go to step 4.2;

[0065] Figure 5 It is a control block diagram of a multi-objective intelligent optimization and efficient control method for a cascaded multi-level electric drive system;

[0066] Among them, 5.1 is the speed closed-loop PI regulator, 5.2 is the coordinate transformation module, 5.3 is the carrier phase shift modulation strategy, 5.4 is the cascaded multi-level converter, 5.5 is the particle swarm optimization algorithm, 5.6 is the multi-objective model of motor torque ripple and electric drive system efficiency, and 5.7 is the three-phase permanent magnet synchronous motor;

[0067] Figure 6 It is a surface diagram of the effect of magnetizing current and switching frequency on the loss of the electric drive system under fixed working conditions;

[0068] Among them, the working point where the particle swarm algorithm converges is marked on the loss curve. It can be clearly seen from the figure that the particle swarm algorithm converges to the global optimal point;

[0069] Figure 7 The multi-objective intelligent optimization and efficient control method of the cascaded multi-level electric drive system is used in the experiment, and the efficiency diagram of the electric drive system in the entire working range is shown;

[0070] Among them, (a) is the efficiency diagram of the electric drive system with zero d-axis current control; (b) is the efficiency diagram of the electric drive system after adopting particle swarm intelligent optimization control;

[0071] Figure 8 This is the waveform diagram before and after torque ripple suppression when zero d-axis current control is used at a speed of 30 rpm.

[0072] Among them, (a) is the actual current and torque waveform before harmonic current suppression is injected, (b) is the actual current and torque spectrum before harmonic current suppression is injected, (c) is the actual current and torque waveform after harmonic current suppression is injected, and (d) is the actual current and torque spectrum after harmonic current suppression is injected;

[0073] Figure 9 This is a diagram showing the optimization effect of the multi-objective intelligent optimization and efficient control method for the cascaded multi-level electric drive system on the motor torque ripple and electric drive system loss under the rated speed and three load torque conditions of 0.6pu, 0.8pu, and 1pu.

[0074] Among them, (a) is the loss histogram of the electric drive system after zero d-axis current control and particle swarm control. It can be seen that the loss of the particle swarm algorithm is significantly reduced under the three working conditions; (b) is the motor torque pulsation value under the three working conditions after zero d-axis current control and particle swarm optimization. The particle swarm algorithm also has a significant suppression effect on torque pulsation, which shows that the particle swarm algorithm achieves multi-objective optimization control effect. DETAILED DESCRIPTION

[0075] The inventive concept of the present invention is further illustrated below with reference to the accompanying drawings and specific embodiments. It should be understood that these embodiments are only used to illustrate the present invention and are not used to limit the scope of the present invention. After reading the present invention, various equivalent modifications made by those skilled in the art to the present invention all fall within the scope defined by the claims attached to the present invention.

[0076] The equivalent circuit diagram of the permanent magnet synchronous motor model without considering the wide frequency domain loss of the motor according to the present invention is as follows: Figure 1As shown, the stator resistance 1.1, the q-axis coupled potential 1.2, the d-axis stator voltage 1.3, the d-axis inductance 1.4, the d-axis coupled potential 1.5, the q-axis inductance 1.6, and the q-axis stator voltage 1.7 are included. The d-axis and q-axis equivalent circuits are composed of resistance, inductance, and potential in series. Based on the equivalent circuit structure, the following motor model can be obtained:

[0077]

[0078] Among them, u d and u q are the d-axis and q-axis stator voltages, i d and i q are the d-axis and q-axis stator currents, L d and L q are the d-axis and q-axis inductances, R s is the stator resistance, ψ f is the permanent magnet flux linkage, and ω is the electrical angular frequency.

[0079] The modeling process of the motor wide-frequency domain loss model used in the present invention to optimize the control strategy assumes the following:

[0080] (1) Ignore spatial harmonics and assume that the winding magnetomotive force is distributed sinusoidally along the air gap;

[0081] (2) Ignore the magnetic circuit saturation effect, that is, the self-inductance of each winding and the mutual inductance between windings are fixed;

[0082] (3) Ignore the influence of temperature rise and frequency on the motor parameters;

[0083] (4) Assume that the internal magnetic permeability of the permanent magnet is the same as that of air, that is, the value of the permanent magnet flux is constant

[0084] (5) High-frequency PWM excitation is mainly the switching frequency and its surrounding frequency band, which is much higher than the fundamental frequency. Therefore, the motor winding phase resistance R can be ignored when calculating the motor winding impedance. s The influence of the inductance is considered only

[0085] The equivalent circuit diagram of the permanent magnet motor taking into account the wide frequency domain loss model of the motor described in the present invention is as follows: Figure 2 As shown, 2.1 is the stator winding resistance, 2.2 is the coupling potential of the q-axis to the d-axis, 2.3 is the d-axis voltage applied to the stator when the converter is powered, 2.4 is the equivalent resistance of the motor's wide-band loss, 2.5 is the d-axis inductance, 2.6 is the coupling potential of the d-axis to the q-axis, 2.7 is the q-axis voltage applied to the stator when the converter is powered, and 2.8 is the q-axis inductance.

[0086] This model directly connects an equivalent iron loss resistance of 2.4 in parallel to the inductor and potential branch (i.e., the induced potential branch). The equivalent iron loss resistance in parallel usually requires a large number of experimental tests and fitting to obtain. Based on the equivalent circuit, the motor model shown in formula (2) can be obtained:

[0087]

[0088] Among them, u df and u qf ,u dh and u qh are the dq axis fundamental frequency voltage and dq axis high frequency voltage respectively, i df and i qf ,i dh and i qh are the dq axis fundamental frequency current and dq axis high frequency current respectively, p is the derivative symbol, L d , L q are the dq axis inductance, ψ p is the permanent magnet flux, ω e is the electrical angular frequency; R i Is the equivalent iron loss resistance, which represents the result of the combined effect of the equivalent iron loss resistance of each frequency. For the cascaded multi-level converter using carrier phase shifting, the main harmonics of the output voltage are concentrated at 1 times the equivalent switching frequency and 2 times the equivalent switching frequency. Therefore, R i Including the fundamental frequency iron loss equivalent resistance R c , iron loss equivalent resistance R at 1 times equivalent switching frequency f1 , iron loss equivalent resistance R at 2 times the equivalent switching frequency f2 Based on the generation mechanism of the conduction loss and the switching loss of the converter, the converter loss model proposed in the present invention is shown as follows:

[0089]

[0090] Among them, V on is the collector-emitter voltage when conducting, R on is the on-resistance, e on ,e off ,e rec are the reference values ​​of turn-on loss, turn-off loss and reverse recovery loss, respectively. c is the switching frequency, in this embodiment V on ,R on ,e on ,e off ,e rec All of them are standard experimental values ​​in the device manual. In the above formula, the current parameter I mis the three-phase current amplitude. According to the Park transform, the three-phase current amplitude is the square root of the sum of the d-axis and q-axis currents. Therefore, the converter loss model and the motor wide-frequency domain loss model have the same independent variable: d-axis current. The present invention can effectively integrate the converter loss model and the motor wide-frequency domain loss model, making it easier to apply the optimization algorithm. Therefore, the loss model for the cascaded multi-level electric drive system is as follows (4):

[0091] P L =minP Cu +P Fe,f (i od )+P Fe,h (i od )+P inv (i od ) (4)

[0092] Among them, P L is the electric drive system loss model, P Cu is the copper loss, P Fe,f is the motor fundamental frequency loss, P Fe,h is the high-frequency loss of the motor, P inv is the converter loss.

[0093] The wide-band loss model for electric motors shows that the magnetizing current trajectory directly affects the calculated copper losses and fundamental frequency iron losses. This current trajectory, in turn, affects the modulation index, which indirectly influences the harmonic voltage output of the converter, thus affecting high-frequency losses. From the converter's perspective, the current trajectory influences the current amplitude, causing changes in the converter's conduction losses, while the frequency directly affects the converter's switching losses. Therefore, optimizing the current trajectory is crucial for optimizing losses in cascaded multilevel electric drive systems.

[0094] The present invention provides a multi-objective intelligent optimization and efficient control method for a cascaded multi-level electric drive system, the flow chart of which is as follows: Figure 3 As shown, it includes step 3.1 of establishing an electric drive system loss model, step 3.2 of proposing an electric drive loss optimization control method, step 3.3 of proposing a permanent magnet synchronous motor torque pulsation model, step 3.4 of proposing a torque pulsation suppression strategy based on harmonic injection, step 3.5 of constructing an electric drive loss model that takes torque pulsation into account, step 3.6 of proposing a particle swarm optimization algorithm, and step 3.7 of proposing a multi-objective and multi-dimensional optimization algorithm for an electric drive system.

[0095] A calculation example of step 3.2 of the particle swarm optimization model based on the current angle variable of the cascaded multi-level motor drive system proposed in the present invention is shown in the following formula:

[0096]

[0097] Among them, i odis the magnetizing current, representing the current angle, T e The torque formula constrains the trajectory of the current under the same working conditions, p n is the pole pair number, ψ p is the permanent magnet flux linkage, and the inequality constraints are the current constraint conditions and voltage constraint conditions, which ensure that the motor operates within a safe range. sm The maximum voltage for safe operation of the motor, i sm L is the maximum current value for safe operation of the motor. d and L q are the d-axis and q-axis inductances respectively; ω e is the electrical angular frequency; i s is the stator current, i d i q is the dq axis current, i sm The maximum current value for safe operation of the motor; i od,min is the minimum value of magnetizing current, i od,max is the maximum value of magnetizing current; f sw is the switching frequency, f sw,min is the minimum switching frequency for normal operation of the motor, f sw,max The maximum switching frequency for normal motor operation;

[0098] The torque pulsation model proposed in the present invention is step 3.3. Generally, according to the magnetic field modulation theory, the electromagnetic torque is the effect of the air gap magnetic energy on the rotor position. Derivative:

[0099]

[0100] Among them, P is the constant term composed of the motor structure and electromagnetic constant, F f is the permanent magnet potential, F a is the armature magnetic potential, θ is the motor space angle, and t is the time. According to the magnetic field modulation theory, the torque generated is divided into three parts: the torque T generated by the rotor magnetic field modulated by the stator slots ff , the torque generated by the interaction between the stator and rotor magnetic fields (permanent magnet torque) T fa , the component T of the armature magnetic field modulated by the stator slots aa , for T aa The size is independent of the rotor position and does not actually contribute to torque. The conditions for generating constant torque are: the coefficients of θ and t are both 0. The conditions for generating torque pulsation are: the coefficient of θ is 0; the coefficient of t is not 0. By analyzing the above constraints, it can be seen that: for three-phase motors, any odd-order current harmonics can generate average torque and torque pulsations of multiples of 6. Therefore, the 5th and 11th order current harmonics hardly generate average torque. According to the established torque model, the relationship between the 6th and 12th order torque pulsations generated by the fundamental phase current and the amplitude and phase of the fundamental phase current is as follows:

[0101]

[0102] Among them, T 1_6 The fundamental current generates 6 times the torque ripple, T 1_12 Generates 12 times of torque pulsation for fundamental current, is the phase of the average torque, Generates 6 torque ripple phases for the fundamental wave, Generate 12 torque ripple phases for the fundamental wave, is the phase difference between the fundamental wave and the 6th torque ripple, K is the phase difference between the fundamental wave and the 12th torque ripple, 1_6 ,K 1_12 are the constants for the 6th and 12th order torque ripples generated by the fundamental current, respectively. I1 is the square root of the motor's dq axis current amplitudes. Finite element analysis shows that the 6th order torque ripple amplitude does not exceed 25% of the 12th order torque ripple amplitude under different operating conditions. Therefore, the 6th order torque ripple is ignored. To unify it with the current variable during control optimization, a quantitative relationship between the 12th order torque ripple and the three-phase current is established as follows:

[0103]

[0104] Among them, I1 is the square root of the motor dq axis current amplitude, is the phase of the average torque.

[0105] The present invention proposes step 3.4 of the torque pulsation suppression strategy based on harmonic current injection. It can be seen from the mathematical expression of the torque model that for currents of a specific frequency and phase, ignoring magnetic circuit saturation, the average torque is proportional to the injected current amplitude. Taking the 12th torque pulsation generated by the 11th current harmonic of a specific amplitude and phase as an example, the 12th torque pulsation generated by the 11th current harmonic and the fundamental wave is considered, and the 12th torque pulsation generated by the fundamental wave alone is subtracted from the 12th torque pulsation, thereby obtaining the accurate amplitude and phase of the 12th torque pulsation generated by the 11th current harmonic. As shown in the following formula:

[0106]

[0107] Where T 1_12 、T 11_12 、T 12 They are the fundamental wave generation, 11th current harmonic generation and the total 12th torque ripple amplitude, They are the phases for fundamental wave generation, 11th current harmonic generation, and the total 12th torque ripple respectively. Therefore, for the above analysis under fixed operating conditions, for the 12th torque ripple generated by the fundamental wave, only an 11th current harmonic needs to be injected to generate a 12th torque ripple with the same amplitude and opposite phase to cancel it. The injected harmonic current required to suppress the torque ripple can be calculated according to the above formula. However, injecting harmonic current into the motor drive system will increase the motor losses at the corresponding frequency. As shown in the following formula:

[0108]

[0109] where P inj_loss is the loss generated by the injected current, R inj_f is the small resistance corresponding to the loss at the injected current frequency, i inj , represents the amplitude of the injected harmonic current, that is, the square root of the sum of the squares of the dq-axis injected currents I d_inj 、I q_inj .

[0110] In step 3.5 of the particle swarm multi-objective optimization algorithm described in the present invention, a penalty function is introduced, and the torque ripple is used as a constraint condition to transform the multi-objective optimization into an optimal operating point that realizes efficiency optimization on the premise of satisfying complex constraint conditions, as shown in the following formula:

[0111] Fitness = P L + P inj_loss + pen(I(C1, T 12 )) (11)

[0112] where Fitness is the multi-objective optimization objective function, P L is the loss of the electric drive system, P inj_loss is the loss of the injected harmonic current, pen is the penalty function, I(x, y) is the indicator function, if x < y, the function outputs 1, if x > y, the function outputs 0. T 12 is the 12th torque ripple model of the motor, C1 is the boundary amplitude of the torque ripple under the torque ripple index, and it is judged whether the torque ripple requirement is met under the operating conditions. The torque ripple performance index during operation is set in advance to constrain the range of particle search. pen is generally set to a relatively large value. When the particle searches to this area, the objective function value is far from meeting the minimum value, and it will transmit information to the entire particle swarm to avoid searching this area. Because injecting harmonic current can significantly reduce the torque ripple amplitude, but it will also increase the loss of the electric drive system. The application of the penalty function ensures that the efficiency of the electric drive system is improved as much as possible under the requirement of suppressing the torque ripple.

[0113] A calculation of step 3.6 of the particle swarm algorithm for the loss model of the cascaded multi-level electric drive system described in the present invention is shown in the following formula:

[0114]

[0115] Among them, x ij is the position of the particle, c1 and c2 are learning factors, also known as acceleration constants; r1 and r2 are uniform random numbers in the range [0, 1], i = 1, 2, ... D; v ij is the particle speed, and is limited to the speed set by the user. r1 and r2 are random numbers between 0 and 1, which increase the randomness of particle flight. ij is the optimal position of the i-th particle, p gj is the optimal position of all particles. The flow chart of particle swarm algorithm is as follows Figure 4 As shown, it includes initializing the particle position and speed 4.1, calculating the value of each particle in the objective function 4.2, calculating the minimum objective function value corresponding to each particle 4.3, calculating the minimum objective function value in the group 4.4, optimizing the particle speed and position 4.5, executing the boundary conditions 4.6, and judging whether the convergence conditions are met 4.7. If so, output the result 4.8, otherwise jump to 4.2.

[0116] Step 3.7 of the particle swarm algorithm for the cascaded multi-level electric drive system loss model of the present invention is shown in the following formula.

[0117]

[0118] The control block diagram of the particle swarm optimization strategy based on the cascade multi-level electric drive system loss model of the present invention is as follows: Figure 5 As shown, the system includes a speed closed-loop PI regulator 5.1, a coordinate transformation module 5.2, a carrier phase-shift modulation strategy 5.3, a cascaded multilevel converter 5.4, a particle swarm optimization algorithm 5.5, a multi-objective model for motor torque ripple and electric drive system efficiency 5.6, and a three-phase permanent magnet synchronous motor 5.7. The encoder, torque analyzer, and current sensor collect motor speed, torque ripple components, and three-phase currents. The three-phase currents are then transformed using Park transforms to obtain the d / q axis currents, which are then input into the multi-objective model. The particle swarm multi-objective algorithm is then applied to solve for the optimal injection current and d / q axis current trajectories. The speed loop PI is tuned to determine the rated stator current amplitude. Based on the torque equation constraints, the d / q axis rated values ​​are calculated separately and then output through current loop tuning.

[0119] The curve diagram of the current trajectory of the present invention for the change of electric drive loss is as follows: Figure 6 As shown in the figure, the change of current change to loss under fixed working conditions is plotted. The marked points in the figure are the working points where the particle swarm search algorithm converges, which shows the effectiveness of particle swarm optimization for the electric drive loss model.

[0120] The present invention applies experimental verification to verify the effect of the proposed particle swarm optimization algorithm. Figure 7 shown. Figure 7 It is based on the loss model of cascaded multi-level electric drive system. In the full working range, the application of i d =0 and the efficiency map of the two control strategies of particle swarm optimization.

[0121] The experimental results of the torque ripple suppression strategy of the present invention are as follows: Figure 8 As shown in the figure, the operation is carried out at a speed of 30 rpm, and zero d-axis current control is adopted. (a) and (c) are the actual current and torque waveforms and torque spectra before harmonic current suppression is injected; (b) and (d) are the actual current and torque waveforms and torque spectra after harmonic current suppression is injected.

[0122] The results of the simulation verification using Simulink and Ansys are as follows: Figure 9 , which is the optimization effect diagram of the torque pulsation and motor drive system loss under the rated speed and load conditions of 0.6pu, 0.8pu, and 1pu after applying the particle swarm multi-objective optimization function. It can be seen that under the three conditions, the motor loss is significantly reduced, the 12th torque pulsation amplitude is also reduced, and the torque performance is improved.

[0123] The parameters of the permanent magnet motor used in the experiment are as follows: 10 pole pairs, 12.773mH q-axis inductance, 8.268mH d-axis inductance, 0.10397Wb permanent magnet flux linkage, 0.56Ω stator resistance, and a sampling frequency of 2kHz. The proposed particle swarm optimization strategy for permanent magnet synchronous motors based on a wide-band loss model for the motor, compared to conventional optimization algorithms, has superior global search capabilities, and the effectiveness of the optimization can be seen in the experimental optimization range.

[0124] It will be understood by those skilled in the art that the embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. The solutions in the embodiments of the present invention may be implemented in various computer languages, for example, the object-oriented programming language Java and the interpreted scripting language JavaScript.

[0125] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowcharts and / or block diagrams, as well as combinations of processes and / or blocks in the flowcharts and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowcharts and / or block diagrams. Figure 1 a process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0126] These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0127] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing the instructions executed on the computer or other programmable device for implementing the process. Figure 1 a process or multiple processes and / or boxes Figure 1 A step that specifies a function in one or more boxes.

[0128] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0129] Obviously, those skilled in the art may make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if such changes and modifications fall within the scope of the claims and their equivalents, the present invention is intended to include such changes and modifications.

Claims

1. A multi-objective intelligent optimization and efficient control method for a cascaded multi-level electric drive system, characterized in that The steps include: Step 1: Construct a wide-frequency domain loss model for the motor and a converter loss model, unify the current variables, and obtain the electric drive system loss model; Step 2: Based on fundamental current angle optimization, a cascaded multi-level permanent magnet synchronous electric drive loss optimization model is designed; Step 3: Establish a permanent magnet synchronous motor torque ripple model based on the magnetic field modulation principle, obtain a mathematical mapping relationship between current harmonic components and torque ripple components, and derive a motor torque ripple model with specific sub-current harmonic components as components; Step 4: Based on the motor torque ripple model, design a motor torque ripple suppression method based on injection of specific sub-current harmonic components; Step 5: Based on the electric drive system loss model constructed in step 1 and the motor torque ripple model obtained in step 3, a multi-objective optimization model for electric drive system loss is established, which takes the fundamental current angle and specific subharmonic current components as two-dimensional optimization variables and takes into account torque ripple performance. Step 6: Based on the multi-objective optimization model of electric drive system loss that takes into account torque ripple performance obtained in Step 5, a particle swarm algorithm is applied to perform particle optimization configuration in the two-dimensional optimization variables of the fundamental current angle and the specific subharmonic current component; In step 7, based on the particle optimization configuration in step 6, the multi-objective optimization model of the electric drive system loss established in step 5 is solved to obtain the optimal fundamental current angle and optimal harmonic current component settings of the electric drive system.

2. The method according to claim 1, wherein: In step 1, the expression of the electric drive system loss model is: P L =minP Cu +P Fe,f (i od )+P Fe,h (i od )+P inv (i od ) Among them, P L is the electric drive system loss, P Cu is the copper loss, P Fe,f is the motor fundamental frequency loss, P Fe,h is the high-frequency loss of the motor, P inv is the converter loss.

3. The method according to claim 1, wherein: In step 1, the motor wide-frequency domain loss model is constructed as follows: Among them, u df and u qf ,u dh and u qh are the dq axis fundamental frequency voltage and dq axis high frequency voltage respectively, R s is the stator resistance; i df and i qf ,i dh and i qh are the dq axis fundamental frequency current and dq axis high frequency current respectively, p is the derivative symbol, L d and L q are the dq axis inductance, ψ p is the permanent magnet flux, ω e is the electrical angular frequency; R i is the equivalent iron loss resistance; Among them, the equivalent iron loss resistance R i Including the fundamental frequency iron loss equivalent resistance R c , iron loss equivalent resistance R at 1 times equivalent switching frequency f1 and the equivalent resistance R of the iron loss at 2 times the equivalent switching frequency f2 , the converter loss model is constructed as, Among them, P inv is the converter loss, V on is the collector-emitter voltage when conducting, R on is the on-resistance, I m is the three-phase current amplitude; e on ,e off ,e rec are the reference values ​​of turn-on loss, turn-off loss and reverse recovery loss, respectively. c is the switching frequency, V dc is the DC side voltage of the H-bridge.

4. The method according to claim 1, wherein: In step 2, the cascaded multi-level permanent magnet synchronous electric drive loss optimization model is constructed as follows: Optimization target P L =minP Cu +P Fe,f (i od )+P Fe,h (i od )+P inv (i od ) Equality constraint T e =1.5p n i oq [ψ p +(L d -L q )i od ] Inequality constraints Among them, P L is the electric drive system loss, P Cu is the copper loss, P Fe,f is the motor fundamental frequency loss, P Fe,h is the high-frequency loss of the motor, P inv is the converter loss, i od is the magnetizing current; T e is the torque, p n is the pole pair number, ψ p is the permanent magnet flux, L d and L q are the d-axis and q-axis inductances respectively; U sm is the maximum voltage amplitude for safe operation of the motor, ω e is the electrical angular frequency; i s is the stator current, i d ,i q are dq axis current, i sm The maximum current value for safe operation of the motor; i od,min is the minimum value of magnetizing current, i od,max is the maximum value of magnetizing current; f sw is the switching frequency, f sw,min is the minimum switching frequency for normal operation of the motor, f sw,max is the maximum switching frequency for normal operation of the motor; its optimization method is to control the fundamental current angle to obtain the minimum value of the objective function.

5. The method according to claim 1, wherein: In step 3, the permanent magnet synchronous motor torque ripple model is obtained based on the magnetic field modulation principle as follows: Among them, P is the constant term composed of the motor structure and electromagnetic constant, F f is the permanent magnet potential, F a is the armature magnetic potential, θ is the motor space angle, and t is the time; according to the magnetic field modulation theory, the torque generated is divided into three parts: the torque T generated by the rotor magnetic field modulated by the stator slots ff , the torque T generated by the interaction between the stator and rotor magnetic fields fa , the component T of the armature magnetic field modulated by the stator slots aa , for T aa Its size is independent of the rotor position and does not contribute to torque.

6. The method according to claim 1, wherein: In step 3, the mathematical mapping relationship between the current harmonic component and the torque ripple component is obtained, and the motor torque ripple model with the specific sub-current harmonic component as the component is derived as follows: Among them, T 12 is the total 12 torque pulsation amplitudes, ω is the electrical angle, Generates 12 torque ripple phases for the fundamental wave; K 1_12 is the constant of torque pulsation generated by fundamental current, I1 is the square root sum of the motor dq axis current amplitude, is the fundamental torque phase, It is the phase difference between the fundamental wave and the 12th order torque ripple.

7. The method according to claim 1, wherein: In step 4, the expression for suppressing the 12th torque ripple based on harmonic current injection is: Among them, T 1_12 、T 11_12 、T 12 They are the fundamental wave generation, 11th current harmonic generation and the total 12th torque ripple amplitude, They are respectively the fundamental wave generation, the 11th current harmonic generation and the total 12th torque ripple phase, and ω is the electrical angle.

8. The method according to claim 1, wherein: In step 5, the penalty function pen is introduced to construct a multi-objective optimization model for electric drive system loss: Optimization target Fitness = minP Cu +P Fe,f (i od )+P Fe,h (i od )+P inj_loss (i inj )+pen(I(C1,T 12 )) Equality constraint T e =1.5p n i oq [ψ p +(L d -L q )i od ] Inequality constraints Among them, Fitness is the multi-objective optimization objective function, P Cu is the motor copper loss P Cu , P Fe_f is the fundamental frequency iron loss, P Fe_h is the high frequency iron loss, P inj_loss is the loss caused by the injected current harmonics, pen is the penalty function, C1 is a constant; I is the discriminant function, if the 12th order torque ripple T 12 If the torque pulsation amplitude is greater than the torque pulsation boundary amplitude C1, the I output is 1, otherwise it is 0.

9. The method according to claim 1, wherein: In step 6, particle optimization configuration is performed in the two-dimensional optimization variables of the fundamental current angle and the specific subharmonic current component as follows: v n (i' d )=wv n (i d )+c1r1[p n (i d )-x n (i d )]+c2r2[g best -x n (i d )] x n (i’ d )=x n (i d )+v n (i' d ) v n (i’ 11_mag )=wv n (i 11_mag )+c1r1[p n (i 11_mag )-x n (i 11_mag )]+c2r2[g best -x n (i 11_mag )] x n (i’ 11_mag )=x n (i 11_mag )+v n (i 11_mag ) v n (i’ 11_phase )=wv n (i 11_phase )+c1r1[p n (i 11_phase )-x n (i 11_phase )]+c2r2[g best -x n (i 11_phase )] x n (i’ 11_phase )=x n (i 11_phase )+v n (i 11_phase ) Among them, v n (i' d ) is the updated velocity of the particle in the d-axis current dimension, p n (i d ) is the optimal position of the nth particle under the d-axis current dimension, g best is the optimal position of the group under the d-axis current dimension, v n (i' 11_mag ) is the updated velocity of the particle in the 11th current harmonic amplitude dimension, p n (i 11_mag ) is the optimal position of the nth particle in the 11th harmonic amplitude dimension, x n (i' 11_mag ) is the updated position of the particle in the 11th current harmonic amplitude dimension, v n (i' 11_phase ) is the updated velocity of the particle in the 11th current harmonic angle dimension, p n (i 11_phase ) is the optimal position of the nth particle in the 11th harmonic angle dimension, x n (i' 11_phase ) is the updated position of the particle in the 11th current harmonic angle dimension; v n (i d ),x n (i d ) are the velocity and position of the nth particle in the d-axis current dimension; v n (i 11_mag ),x n (i 11_mag ) are the velocity and position of the nth particle under the dimension of the amplitude of the injected 11th current harmonic; v n (i 11_phase ),x n (i 11_phase ) are the velocity and position of the nth particle under the phase dimension of the injected 11th current harmonic; w is the inertia weight, c1 and c2 are learning factors; r1 and r2 are uniform random numbers in the range of [0, 1].

10. The method according to claim 1, wherein: In step 7, multiple variables of the multi-objective optimization model are updated and positioned in the multi-dimensional space, thereby searching for the global optimal point, that is, the optimal operating current and the optimal injection current of the electric drive system.

Images