Permanent magnet synchronous motor efficiency optimization method based on optimal system efficiency

Abstract

The invention provides a permanent magnet synchronous motor efficiency optimization method based on optimal system efficiency, which comprises the following steps of: S1, analyzing a loss mechanism and a dynamic efficiency characteristic of each component of an electric drive system of an electric commercial vehicle to obtain each component of the electric drive system and an integral dynamic efficiency model of the electric drive system under different working conditions; s2, formulating a table lookup method flux weakening control strategy of the permanent magnet synchronous motor based on experimental data to serve as a basis of a subsequent permanent magnet synchronous motor control strategy based on optimal system efficiency; and S3, introducing an electric drive system overall dynamic efficiency model and a depth deterministic strategy gradient algorithm to improve look-up table flux-weakening control, obtaining a permanent magnet synchronous motor flux-weakening control strategy based on optimal system efficiency, and realizing efficiency optimization of the permanent magnet synchronous motor. Compared with a traditional motor look-up table flux weakening control strategy, decoupling control over the permanent magnet synchronous motor can be achieved, efficiency optimization of the permanent magnet synchronous motor can be conducted according to the dynamic efficiency characteristic of an electric drive system, and the working efficiency of the vehicle permanent magnet synchronous motor is improved.

Description

Technical Field

[0001] The invention relates to a method for optimizing the efficiency of a permanent magnet synchronous motor, and in particular to a method for optimizing the efficiency of a permanent magnet synchronous motor with optimal system efficiency. Background Art

[0002] Permanent magnet synchronous motors have been widely used in the new energy vehicle industry due to their advantages of small size and high power density. However, permanent magnet synchronous motors have the characteristics of nonlinearity and strong coupling. How to use advanced control strategies to achieve a reasonable improvement in the efficiency of permanent magnet synchronous motors has always been a difficult link to control.

[0003] At present, the efficiency optimization control technology of permanent magnet synchronous motor is divided into the minimum loss control strategy (Loss Minimization Control, LMC) and the search control strategy (Search Control, SC). The minimum loss control strategy uses the objective function of minimum loss and establishes the motor loss model through accurate electromagnetic parameters to solve the optimal control variables such as id and iq to control the motor. The implementation of this type of control strategy is more dependent on the accuracy of motor parameter identification. Changes in the electromagnetic parameters of the motor during the control process will cause the inability to achieve AC-DC current decoupling control under some working conditions. The search control strategy calculates the real-time efficiency of the system by online detection of the state variables such as the input and output power of the motor system, and then adjusts the control parameters so that the motor continues to work at the highest efficiency point. This type of optimization strategy targets the state variables of the motor itself, ignores the influence of other key components of the electric drive system, and cannot achieve optimal control of the electric drive system.

[0004] In summary, in order to achieve optimal control of the efficiency of new energy vehicle drive system and maximize the social value of new energy vehicles, based on the limitations of the current permanent magnet synchronous motor minimum loss control strategy and search control strategy, this paper proposes a permanent magnet synchronous motor efficiency optimization method based on system efficiency optimization to solve the problems that the existing technology has not achieved true decoupling control and the efficiency of the electric drive system cannot reach the optimal level. Summary of the invention

[0005] In view of the shortcomings of the existing technology, the present invention proposes a permanent magnet synchronous motor efficiency optimization method based on optimal system efficiency, which helps to solve the problem that the existing control strategy optimization scheme cannot achieve the best efficiency during the overall operation of the electric drive system.

[0006] The technical solution adopted by the present invention is:

[0007] A method for optimizing the efficiency of a permanent magnet synchronous motor based on optimal system efficiency comprises the following steps:

[0008] Step S1: Based on the electric drive system structure of a pure electric commercial vehicle, analyze the loss mechanism and dynamic efficiency characteristics of the electric drive system, and formulate efficiency models of key components and efficiency models of the electric drive system under different working conditions;

[0009] Step S2: formulating a permanent magnet synchronous motor weak magnetic field control strategy using a table lookup method based on experimental data as a basis for a subsequent permanent magnet synchronous motor control strategy based on optimal system efficiency;

[0010] Step S3: The overall dynamic efficiency model of the electric drive system and the deep deterministic strategy gradient algorithm are introduced to improve the table lookup method of weak magnetic field control, and the permanent magnet synchronous motor weak magnetic field control strategy based on the optimal system efficiency is obtained to achieve the efficiency optimization of the permanent magnet synchronous motor.

[0011] Preferably, the specific process of step S1 is as follows:

[0012] Step S11: Based on the electric drive system structure of the electric commercial vehicle, a dynamic efficiency and loss model of the electrical system including the battery, converter, inverter, motor, air conditioner and the mechanical system is obtained through theoretical analysis;

[0013] Step S12: establishing electric drive system dynamic efficiency models for different vehicle driving conditions: 1) driving condition, 2) air conditioning operating condition.

[0014] Preferably, the specific process in step S2 is as follows:

[0015] Step S21: obtaining key parameters of the permanent magnet synchronous motor and a dq axis current table according to experiments;

[0016] Step S22: Establishing a permanent magnet synchronous motor weak magnetic field control strategy using a table lookup method.

[0017] Preferably, the specific process in step S3 is as follows:

[0018] Step S31: Introduce the overall dynamic efficiency model of the electric drive system and the deep deterministic policy gradient algorithm, establish the state space, action space, and reward function based on the vehicle state information, and form the FW-DDPG control strategy. Its framework is as follows: Figure 5 As shown;

[0019] Step S32: The FW-DDPG algorithm is trained using the reconstructed working condition, and after the algorithm converges, a MAP diagram of the permanent magnet synchronous motor under the control of multiple different control strategies is obtained.

[0020] Beneficial effects of the invention:

[0021] 1. The present invention is based on a permanent magnet synchronous motor efficiency optimization method with optimal system efficiency. By introducing deep reinforcement learning and an electric drive system efficiency model, it realizes the decoupling control of the permanent magnet synchronous motor in a true sense and improves the working efficiency of the automotive permanent magnet synchronous motor.

[0022] 2. Compared with the traditional motor lookup table method weak magnetic control strategy, this application does not rely on the parameters identified by the motor. It solves the problem that the AC and DC axis current decoupling control cannot be achieved under some working conditions due to changes in the motor electromagnetic parameters during the control process, and realizes true decoupling control.

[0023] 3. This application breaks the inertial thinking of traditional motor control that only considers its own state variables, takes into account the influence of other key components of the electric drive system, and introduces the efficiency characteristics of the electric drive system under different working conditions as the state variables of the motor control into the DDPG algorithm to form a FW-DDPG control strategy. This strategy can control the permanent magnet synchronous motor according to the actual working conditions of the vehicle to optimize the efficiency of the permanent magnet synchronous motor.

[0024] 4. The efficiency optimization method of the permanent magnet synchronous motor based on the optimal vehicle efficiency is proposed based on pure electric commercial vehicles, but it still has certain applicability to the efficiency optimization of permanent magnet synchronous motors of other models. BRIEF DESCRIPTION OF THE DRAWINGS

[0025] Figure 1 The figure is a flow chart of the method for optimizing the efficiency of a permanent magnet synchronous motor based on optimal system efficiency according to the present invention;

[0026] Figure 2 Shown is the dynamic efficiency characteristic diagram of the electric drive system;

[0027] Among them, (a) efficiency MAP (driving condition), (b) loss MAP (driving condition), (c) efficiency MAP (air conditioning start condition), (d) loss MAP (air conditioning start condition);

[0028] Figure 3 (a) and (b) are the d-axis current diagram and q-axis current diagram obtained from the experiment respectively;

[0029] Figure 4 The figure shows the principle diagram of the table lookup method field weakening control strategy of the present invention;

[0030] Figure 5 The figure shows the principle diagram of the FW-DDPG control strategy of the present invention;

[0031] Figure 6 The figure shows the motor efficiency MAP comparison diagram between the FW-DDPG control strategy of the present invention and the table lookup weak magnetic control strategy, where (a) is the MAP before optimization and (b) is the MAP after optimization. DETAILED DESCRIPTION

[0032] In order to make the technical concept and advantages of the present invention more clearly understood, the technical solution of the present invention is further described in detail below in conjunction with the accompanying drawings. It should be understood that the following embodiments are only used to explain and illustrate the preferred implementation methods of the present invention, and should not be regarded as and do not constitute a limitation on the scope of patent protection claimed by the present invention.

[0033] Example 1

[0034] See also Figure 1 The present invention provides a method for optimizing the efficiency of a permanent magnet synchronous motor based on the optimal vehicle efficiency, comprising the following steps:

[0035] Step S1: analyzing the loss mechanism and dynamic efficiency characteristics of each component of the electric drive system of the electric commercial vehicle, and obtaining the dynamic efficiency model of each component of the electric drive system and the overall electric drive system under different working conditions;

[0036] Step S2: formulating a permanent magnet synchronous motor weak magnetic field control strategy using a table lookup method based on experimental data as a basis for a subsequent permanent magnet synchronous motor control strategy based on optimal system efficiency;

[0037] Step S3: The overall dynamic efficiency model of the electric drive system and the deep deterministic strategy gradient algorithm are introduced to improve the table lookup method of weak magnetic field control, and the permanent magnet synchronous motor weak magnetic field control strategy based on the optimal system efficiency is obtained to achieve the efficiency optimization of the permanent magnet synchronous motor.

[0038] Example 2

[0039] The difference between this embodiment and embodiment 1 is that:

[0040] The specific process of step S1 is as follows:

[0041] Step S11: Based on the electric drive system structure of the electric commercial vehicle, a dynamic efficiency and loss model of the electrical system including the battery, converter, inverter, motor, air conditioner and the mechanical system is obtained through theoretical analysis;

[0042] Step S12: establishing electric drive system dynamic efficiency models for different vehicle driving conditions: 1) driving condition, 2) air conditioning operating condition.

[0043] The specific process of step S2 is as follows:

[0044] Step S21: obtaining key parameters of the permanent magnet synchronous motor and a dq axis current table according to experiments;

[0045] The motor modeling parameters mainly include stator resistance, d-axis inductance, q-axis inductance, permanent magnet flux, etc. The parameters required for modeling come from two sources: 1) provided by the supplier, 2) parameter identification through experiments;

[0046] Step S22: establishing a permanent magnet synchronous motor weak magnetic control strategy using a table lookup method;

[0047] The table-based weak magnetic control is evolved from the FOC framework. While retaining SVPWM, Clark, and Park, the current loop and speed loop are improved. The current loop, SVPWM, Clark, Park and other modules remain unchanged.

[0048] The specific process of step S3 is as follows:

[0049] Step S31: Introduce the overall dynamic efficiency model of the electric drive system and the deep deterministic policy gradient algorithm, establish the state space, action space, and reward function based on the vehicle state information, and form the FW-DDPG control strategy. Its framework is as follows: Figure 5 As shown;

[0050] The FW-DDPG control strategy uses the table-based weak magnetic control as the basic framework, and uses the DDPG algorithm to replace the FFCDC part to achieve true decoupling control of the permanent magnet synchronous motor. In addition, the efficiency characteristics of the electric drive system under different working conditions are used as part of the DDPG algorithm, so that the control strategy can control the permanent magnet synchronous motor according to the actual working conditions of the vehicle;

[0051] Step S32: The FW-DDPG algorithm is trained using the reconstructed working condition, and a MAP diagram of the permanent magnet synchronous motor under the FW-DDPG control strategy is obtained after the algorithm converges.

[0052] In step S32, the CLTC condition is used as the training condition for the FW-DDPG control strategy, and the decision method is trained and simulated: in order to fully train the FW-DDPG control strategy and shorten the algorithm training time, the motor characteristics are divided into several specific operating points (for example: torque 200Nm, speed 3000rpm) in the motor external characteristic curve for training; after the training is completed, the permanent magnet synchronous motor MAP diagram under the FW-DDPG control strategy is obtained. The comparison of the MAP diagram formed under the table lookup method weak magnetic control and the FW-DDPG control is shown in the figure. Figure 6 shown.

[0053] Example 3

[0054] See also Figure 1 The present invention is based on a permanent magnet synchronous motor efficiency optimization method with optimal vehicle efficiency, and the specific implementation process includes:

[0055] Step S1: Taking a certain pure electric commercial vehicle as an example, based on its electric drive system structure, analyze the loss mechanism and dynamic efficiency characteristics of the electric drive system, and formulate efficiency models of key components and efficiency models of the electric drive system under different working conditions;

[0056] Step S2: constructing a permanent magnet synchronous motor weak magnetic control strategy using a table lookup method;

[0057] Step S3: Introduce the electric drive system efficiency model and the reinforcement learning DDPG algorithm to form a FW-DDPG control strategy to achieve permanent magnet synchronous motor efficiency optimization.

[0058] Specifically, the implementation steps of step S1 are as follows:

[0059] Step S11, the dynamic efficiency characteristics of the electric drive system are analyzed as follows:

[0060] The loss mechanism and dynamic efficiency characteristics of each component of the electric drive system are different. Among them, the power loss of the battery is mainly caused by the charging and discharging current, internal resistance and RC circuit voltage drop; the inverter loss is mainly caused by the switching device (IGBT and freewheeling diode) opening and closing; the converter loss is caused by the parasitic resistance of the switching device, capacitor and inductor; the permanent magnet synchronous motor loss is composed of winding copper loss, stator iron loss, mechanical loss and stray loss; the transmission system loss is composed of gear meshing friction power loss, bearing friction power loss, oil stirring friction power loss and wind resistance power loss; the air conditioning power loss mainly includes transmission loss, volumetric loss and other losses. Establish a mathematical model of the loss and dynamic efficiency characteristics of the electric drive system including batteries, converters, inverters, motors, air conditioners and mechanical transmission systems; Figure 2 This is a dynamic efficiency characteristic diagram of the electric drive system of the present invention.

[0061] Battery loss and efficiency mathematical model:

[0062]

[0063] P RC =NI1(∫((I1-V p1 / R p1 ) / C p1 )+∫((I1-V p2 / R p2 ) / C p2 ));

[0064] η bat =P out / (P out +P R +P RC ); (1)

[0065] Where P R , P RC are the battery internal resistance and polarization circuit power loss, kW; V p1 、V p2 are the first-order polarization voltage and the second-order polarization voltage, V; P p1 , P p1are the first-order polarization resistance and the second-order polarization resistance, Ω; C p1 , C p2 They are the first-order polarization capacitance and the second-order polarization capacitance, F; N is the number of single cells; I1 is the main circuit current of the battery, A.

[0066] Inverter loss and efficiency mathematical model

[0067] P cond_IGBT =I c (V ce0 +I c (R2+K RT ΔT J_IGBT ));

[0068] P cond_Diode =I f (V f0 +I f (R3+K RT' ΔT J_Diode ));

[0069] P sw_IGBT =f(E sw_on +E sw_off );

[0070] P sw_Diode =fE sw_rec ;

[0071] η inverter =P out / (P out +P cond_IGBT +P cond_Diode +P sw_IGBT +P sw_Diode ); (2)

[0072] In the formula, ΔT j_IGBT , ΔT j_Diode is the junction temperature difference between IGBT and Diode, ℃; K RT , K RT' is the on-resistance temperature coefficient of IGBT and Diode, 3.0E-6; f is the switching frequency, Hz; E sw_on 、E sw_off 、E sw_rec IGBT turn-on energy consumption, IGBT turn-off energy consumption, Diode reverse recovery energy consumption, mJ; I c ,I f is the IGBT collector-emitter current, the diode conduction current, A; V ceo 、V fo is the IGBT and Diode threshold voltage, V; R2 and R3 are the IGBT and Diode on-resistance, Ω.

[0073] Converter Loss and Efficiency Model

[0074]

[0075] η dc =P out / (P out +P IGBT +P LC_R ); (3)

[0076] In formula (3), P LC_R , P IGBT is the power loss of inductor parasitic resistance, kW; I2 and I3 are the power loss of capacitor parasitic resistance, kW.

[0077] Motor loss and efficiency model:

[0078]

[0079] P M =B(ω m ) 2 ;

[0080] η pmsm =P out / (P out +P Cu +P Fe +P M ); (4)

[0081] In formula (4), P Cu , P Fe , P M They are the motor iron loss, copper loss and mechanical loss, kW.

[0082] Air conditioning loss and efficiency model

[0083]

[0084] η Hvac =P out / Tn; (5)

[0085] In formula (5), F d 、F c 、F di 、F D They are the heat transfer areas of the roof, body, bottom and power compartment, m2; K d , K c , K di , K D are the heat transfer coefficients of the roof, body, bottom and power compartment respectively; T o , T iare the outside and inside temperatures of the cabin, respectively, K; η is the penetration coefficient of solar radiation through the glass, 0.85; I is the solar radiation intensity on the outer surface of the window, KW / m2; F I is the effective area in the direct direction of the window, m2; ρ s is the absorption coefficient of the outer surface of the car enclosure structure to solar radiation, 0.9; v is the speed of the car, m / s; n is the number of car occupants; V is the amount of fresh air required per person per hour, m3 / (h*person); ρ is the air density, km / m3; h o 、h i is the specific enthalpy value outside and inside the cabin.

[0086] Transmission system loss and efficiency model

[0087]

[0088]

[0089] η mt =(P mesh +P bearing +P churning +P wind ) / T in ω in ; (6)

[0090] In formula (6), P mesh , P bearing , P churning , P wind They are meshing friction power loss, bearing friction power loss, oil stirring friction power loss and windage power loss, kW; T in is the transmission system torque input, Nm; ω in is the transmission system speed input, r / min; f s ,f1,f2,f3,f g They are the tooth surface sliding friction coefficient, the coefficient related to the bearing type, the bearing immersion depth in the lubricating oil, the left bearing seal coefficient, and the right bearing seal coefficient; F nj is the normal load on the contact surface of part j, N; ω 1j ,ω 2j is the angular velocity of the two gears in part j, r / min; r 1j 、r 2j is the combined curvature radius of the two gears in part j, mm; h R is the thickness of the elastic fluid dynamic lubricating oil film, mm; β is the gear helix angle, °; B is the tooth width, mm; F d is the pressure load on the bearing, N; d mis the average radius of the bearing, mm; v is the kinematic viscosity of the lubricating oil at ambient temperature, m2 / s; n is the operating speed of the gear, r / min; D is the gear diameter, mm; L is the length of the gear shaft, mm; A g is the configuration constant; R f 、m t are respectively the roughness coefficient, gear end face modulus, R f =7.93-4.648 / m t ; R is the gear pitch radius, mm; ρ eq is the oil-gas mixture density in the transmission, kg / m3; μ eq is the oil-gas mixture viscosity in the transmission, m2 / s;

[0091] Step S12, the mathematical model of the loss and efficiency of the electric drive system under different working conditions is formulated. The formulated dynamic efficiency characteristics of the electric drive system under different working conditions are as follows: Figure 2 As shown, the formulation process is as follows:

[0092] Mathematical model of electric drive system loss and efficiency under driving conditions

[0093] ΔP sys =ΔP bat +ΔP dc +ΔP inver +ΔP pmsm +ΔP mt ;

[0094] η sys =η bat ·η dc ·η inver ·η pmsm ·η mt ; (7)

[0095] In formula (7), ΔP sys , ΔP bat , ΔP dc , ΔP inver , ΔP pmsm , ΔP mt They are electric drive system power loss, battery power loss, converter power loss, inverter power loss, permanent magnet synchronous motor power loss, and transmission system power loss, kW.

[0096] Mathematical model of loss and efficiency of electric drive system under air conditioning startup condition

[0097]

[0098] In formula (7), is the battery loss including air conditioning system loss, kW; ΔP hvac is the air conditioning system loss, kW;

[0099] The implementation process of step S2 is as follows:

[0100] Step S21: obtaining key parameters of the permanent magnet synchronous motor and a dq axis current table according to experiments;

[0101] Step S22: Establishing a permanent magnet synchronous motor weak magnetic field control strategy using a table lookup method.

[0102] Among them, in step S22, a table lookup method weak magnetic control strategy is established based on experimental data:

[0103] As the basis of this patented control strategy, the table-based weak magnetic control forms a data table of the relationship between magnetic flux, torque and current through measurement data. When the motor is running, the d and q axis current reference values ​​are obtained by table-based weak magnetic control. Because the table-based weak magnetic control does not rely on motor parameters, it effectively avoids the impact of motor parameter changes on control and is widely used in actual engineering.

[0104] Modeling ideas:

[0105] The main parameters are obtained: the d-axis current meter and q-axis current meter obtained through experiments, the stator resistance Rs required for FFCDC, the D-axis inductance, the Q-axis inductance, and the permanent magnet flux f. The d-axis current diagram and q-axis current diagram obtained through experiments are as follows: Figure 3 shown.

[0106] Framework construction: The table-based weak magnetic control is evolved from the FOC framework. While retaining SVPWM, Clark, and Park, the current loop and speed loop are improved. Specifically: the speed loop is used to obtain the reference torque, and then the current table is queried through the speed limit to obtain the d and q axis reference currents as part of the current loop input. The modeling principle of the table-based weak magnetic control strategy of the present invention is as follows: Figure 4 shown.

[0107] The implementation process of step S3 is as follows:

[0108] Step S31: Introduce the overall dynamic efficiency model of the electric drive system and the deep deterministic policy gradient algorithm, establish the state space, action space, and reward function based on the vehicle state information, and form the FW-DDPG control strategy. Its framework is as follows: Figure 5 As shown;

[0109] Step S32: The FW-DDPG algorithm is trained using the reconstructed working condition, and a MAP diagram of the permanent magnet synchronous motor under the FW-DDPG control strategy is obtained after the algorithm converges.

[0110] Among them, in step S31, the FW-DDPG control strategy is formulated as follows:

[0111] The FW-DDPG control strategy uses the table-based weak magnetic control as the basic framework and uses the DDPG algorithm to replace the FFCDC part to achieve true decoupling control of the permanent magnet synchronous motor. In addition, the efficiency characteristics of the electric drive system under different working conditions are used as part of the DDPG algorithm, so that the control strategy can control the permanent magnet synchronous motor according to the actual working conditions of the vehicle.

[0112] Based on the established table-based field weakening control strategy framework, the FW-DDPG control strategy formulation includes three parts: state space, action space, and reward function formulation, which are defined as follows:

[0113] State Space:

[0114]

[0115] In formula (9), n ref , n are the reference speed and actual speed of the motor, rpm; i d_err , i q_err The motor d-axis reference current, the difference between the d-axis reference current and the actual current, the q-axis reference current, the difference between the d-axis reference current and the actual current, A; e ff It is the current efficiency of the electric drive system, 0-1.

[0116] Action Space:

[0117] A=[Δu d ,Δu d ] (10)

[0118] In formula (10), Δu d ,Δu d is the motor d and q axis voltage correction value, -40-40.

[0119] The reward function is:

[0120]

[0121] In formula (11), α, β, γ, χ, ζ are weights; i d_err 、i q_err The difference between the motor d-axis reference current and the actual current, the difference between the d-axis reference current and the actual current, A; e a 、e b is the efficiency value of the electric drive system before and after the motor, 0-1; Δu d ,Δu d is the motor d and q axis voltage correction value, -40-40.

[0122] In step S32, the present application conducts training and simulation analysis on the decision-making method:

[0123] The CLTC condition is used as the training condition for the FW-DDPG control strategy. In order to fully train the FW-DDPG control strategy and shorten the algorithm training time, the motor characteristics are divided into several specific operating points (for example: torque 200Nm, speed 3000rpm) in the motor external characteristic curve for training; after the training is completed, the permanent magnet synchronous motor MAP diagram under the FW-DDPG control strategy is obtained. The MAP diagram formed under the table lookup method weak magnetic control and FW-DDPG control is compared. Figure 6 shown.

[0124] After comparison, it is found that the proportion of high-efficiency area of ​​the FW-DDPG control strategy is greater than that of the table-lookup weak magnetic control, which plays a role in decoupling control and efficiency optimization.

[0125] The present invention is based on a permanent magnet synchronous motor efficiency optimization method with optimal system efficiency, aiming to achieve decoupling control of permanent magnet synchronous motors and improve the working efficiency of automotive permanent magnet synchronous motors. Compared with the traditional motor lookup table method weak magnetic control strategy, this application can achieve decoupling control of permanent magnet synchronous motors and optimize the efficiency of permanent magnet synchronous motors according to the overall dynamic efficiency changes of the electric drive system.

[0126] The above description is only a preferred embodiment of the present invention and does not constitute a limitation of the present invention. Under the guidance of the prior art, those skilled in the art can make other modifications to the implementation of the present invention without creative work. Any modification made within the spirit and principle of the present invention or simple replacement or equivalent replacement, improvement, etc. made by conventional technical means in the field should be included in the protection scope of the present invention.

Claims

1. A method for optimizing the efficiency of a permanent magnet synchronous motor based on optimal system efficiency, characterized in that: The steps include: Step S1: analyzing the loss mechanism and dynamic efficiency characteristics of each component of the electric drive system of the electric commercial vehicle, and obtaining the dynamic efficiency model of each component of the electric drive system and the overall electric drive system under different working conditions; Step S2: formulating a permanent magnet synchronous motor weak magnetic field control strategy using a table lookup method based on experimental data as a basis for a subsequent permanent magnet synchronous motor control strategy based on optimal system efficiency; Step S3: The overall dynamic efficiency model of the electric drive system and the deep deterministic strategy gradient algorithm are introduced to improve the table lookup method of weak magnetic field control, and the permanent magnet synchronous motor weak magnetic field control strategy based on the optimal system efficiency is obtained to achieve the efficiency optimization of the permanent magnet synchronous motor.

2. The method for optimizing the efficiency of a permanent magnet synchronous motor based on optimal system efficiency according to claim 1, characterized in that: The specific process of step S1 is as follows: Step S11: Based on the electric drive system structure of the electric commercial vehicle, a dynamic efficiency and loss model of the electrical system including the battery, converter, inverter, motor, air conditioner and the mechanical system is obtained through theoretical analysis; Step S12: establishing electric drive system dynamic efficiency models for different vehicle driving conditions: 1) driving condition, 2) air conditioning operating condition.

3. The method for optimizing the efficiency of a permanent magnet synchronous motor based on optimal system efficiency according to claim 2, characterized in that: Step S11, the dynamic efficiency characteristics of the electric drive system are analyzed as follows: Establish a mathematical model of the loss and dynamic efficiency characteristics of the electric drive system including batteries, converters, inverters, motors, air conditioners and mechanical transmission systems; Battery loss and efficiency mathematical model: P RC =NI1(∫((I1-V p1 / R p1 ) / C p1 )+∫((I1-V p2 / R p2 ) / C p2 )); the bat =P out / (P out +P R +P RC ); (1) Where P R , P RC are the battery internal resistance and polarization circuit power loss, kW; V p1 、V p2 are the first-order polarization voltage and the second-order polarization voltage, V; P p1 , P p1 are the first-order polarization resistance and the second-order polarization resistance, Ω; C p1 , C p2 are the first-order polarization capacitance and the second-order polarization capacitance, F; N is the number of single cells; I1 is the main circuit current of the battery, A; Inverter loss and efficiency mathematical model P cond_IGBT =I c (V ce0 +I c (R2+K RT ΔT J_IGBT )); P cond_Diode =I f (V f0 +I f (R3+K RT' ΔT J_Diode )); Q sw_IGBT =f(E sw_on +E sw_off ); P sw_Diode =fE sw_rec ; the inverter =P out / (P out +P cond_IGBT +P cond_Diode +P sw_IGBT +P sw_Diode ); (2) In the formula, ΔT j_IGBT , ΔT j_Diode is the junction temperature difference between IGBT and Diode, ℃; K RT , K RT' is the on-resistance temperature coefficient of IGBT and Diode, 3.0E-6; f is the switching frequency, Hz; E sw_on 、E sw_off 、E sw_rec IGBT turn-on energy consumption, IGBT turn-off energy consumption, Diode reverse recovery energy consumption, mJ; I c ,I f is the IGBT collector-emitter current, the diode conduction current, A; V ceo 、V fo is the threshold voltage of IGBT and Diode, V; R2 and R3 are the on-resistance of IGBT and Diode, Ω; Converter Loss and Efficiency Model the dc =P out / (P out +P IGBT +P LC_R ); (3) In formula (3), P LC_R , P IGBT is the inductor parasitic resistance loss power, kW; I2 and I3 are the capacitor parasitic resistance loss power, kW; Motor loss and efficiency model: P M =B(ω m ) 2 ; the pmsm =P out / (P out +P Cu +P Fe +P M ); (4) In formula (4), P Cu , P Fe , P M They are the motor iron loss, copper loss and mechanical loss, kW; Air conditioning loss and efficiency model or Hvac =P out / Tn; (5) In formula (5), F d 、F c 、F di 、F D They are the heat transfer areas of the roof, body, bottom and power compartment, m2; K d , K c , K di , K D are the heat transfer coefficients of the roof, body, bottom and power compartment respectively; T o , T i are the outside and inside temperatures of the cabin, respectively, K; η is the penetration coefficient of solar radiation through the glass, 0.85; I is the solar radiation intensity on the outer surface of the window, KW / m2; F I is the effective area in the direct direction of the window, m2; ρ s is the absorption coefficient of the outer surface of the car enclosure structure to solar radiation, 0.9; v is the speed of the car, m / s; n is the number of car occupants; V is the amount of fresh air required per person per hour, m3 / (h*person); ρ is the air density, km / m3; h o 、h i is the specific enthalpy value outside and inside the cabin; Transmission system loss and efficiency model or mt =(P mesh +P bearing +P churning +P wind ) / T in oh in (6) In formula (6), P mesh , P bearing , P churning , P wind They are meshing friction power loss, bearing friction power loss, oil stirring friction power loss and windage power loss, kW; T in is the transmission system torque input, Nm; ω in is the transmission system speed input, r / min; f s ,f1,f2,f3,f g They are the tooth surface sliding friction coefficient, the coefficient related to the bearing type, the bearing immersion depth in the lubricating oil, the left bearing seal coefficient, and the right bearing seal coefficient; F nj is the normal load on the contact surface of part j, N; ω 1j ,ω 2j is the angular velocity of the two gears in part j, r / min; r 1j 、r 2j is the combined curvature radius of the two gears in part j, mm; h R is the thickness of the elastic fluid dynamic lubricating oil film, mm; β is the gear helix angle, °; B is the tooth width, mm; F d is the pressure load on the bearing, N; d m is the average radius of the bearing, mm; v is the kinematic viscosity of the lubricating oil at ambient temperature, m2 / s; n is the operating speed of the gear, r / min; D is the gear diameter, mm; L is the length of the gear shaft, mm; A g is the configuration constant; R f 、m t are respectively the roughness coefficient, gear end face modulus, R f =7.93-4.648 / m t ; R is the gear pitch radius, mm; ρ eq is the oil-gas mixture density in the transmission, kg / m3; μ eq is the oil-gas mixture viscosity in the transmission, m2 / s.

4. The method for optimizing the efficiency of a permanent magnet synchronous motor based on optimal system efficiency according to claim 3 is characterized in that: Step S12, the loss and efficiency mathematical model of the electric drive system under different working conditions is formulated. The formulated dynamic efficiency characteristics of the electric drive system under different working conditions are shown in Figure 2, and the formulation process is as follows: Mathematical model of electric drive system loss and efficiency under driving conditions ΔP sys =ΔP bat +ΔP dc +ΔP inver +ΔP pmsm +ΔP mt ; or sys =the bat ·or dc ·or inver ·or pmsm ·or mt (7) In formula (7), ΔP sys , ΔP bat , ΔP dc , ΔP inver , ΔP pmsm , ΔP mt They are the power loss of electric drive system, battery power loss, converter power loss, inverter power loss, permanent magnet synchronous motor power loss, and transmission system power loss, kW; Mathematical model of loss and efficiency of electric drive system under air conditioning startup condition In formula (8), is the battery loss including air conditioning system loss, kW; ΔP hvac is the air conditioning system loss, kW.

5. The method for optimizing the efficiency of a permanent magnet synchronous motor based on system optimization according to any one of claims 1 to 4, characterized in that: The specific process of step S2 is as follows: Step S21: obtaining key parameters of the permanent magnet synchronous motor and a dq axis current table according to experiments; Step S22: Establishing a permanent magnet synchronous motor weak magnetic field control strategy using a table lookup method.

6. The method for optimizing the efficiency of a permanent magnet synchronous motor based on optimal system efficiency according to claim 5, characterized in that: In step S22, a table lookup method field weakening control strategy is established based on experimental data: The data table of the relationship between flux linkage, torque and current is formed by measuring the data, which serves as the basis for the weak magnetic field control strategy of the table lookup method. When the motor is running, the reference values ​​of the d and q axis currents are obtained by table lookup to perform weak magnetic field control. The modeling process includes: (1) Obtaining the main parameters: the d-axis ammeter and q-axis ammeter obtained through experiments, the stator resistance Rs required for FFCDC, the D-axis inductance, the Q-axis inductance, and the permanent magnet flux f; (2) Framework construction: Based on the traditional FOC dual closed-loop control logic, in the speed conversion, the speed loop is used to obtain the reference torque, and then the current table is queried through the speed limit to obtain the d and q axis reference currents as part of the current loop input.

7. The method for optimizing the efficiency of a permanent magnet synchronous motor based on optimal system efficiency according to claim 5, characterized in that: The specific process of step S3 is as follows: Step S31: introducing the overall dynamic efficiency model of the electric drive system and the deep deterministic policy gradient algorithm, establishing the state space, action space, and reward function based on the vehicle state information, and forming the FW-DDPG control strategy, the framework of which is shown in FIG5 ; Step S32: The FW-DDPG algorithm is trained using the reconstructed working condition, and a MAP diagram of the permanent magnet synchronous motor under the FW-DDPG control strategy is obtained after the algorithm converges.

8. The method for optimizing the efficiency of a permanent magnet synchronous motor based on optimal vehicle efficiency according to claim 6 is characterized in that: The specific process of step S3 is as follows: Step S31: Introduce the deep deterministic policy gradient algorithm, establish the state space, action space, and reward function based on the vehicle state information, and form the FW-DDPG control strategy; Step S32: The FW-DDPG algorithm is trained using typical working conditions, and after the algorithm converges, a MAP diagram of the permanent magnet synchronous motor under control of multiple different control strategies is obtained.

9. The method for optimizing the efficiency of a permanent magnet synchronous motor based on optimal vehicle efficiency according to claim 6, 7 or 8, characterized in that: In step S31, the FW-DDPG control strategy is formulated as follows: Based on the established table-based field weakening control strategy framework, the FW-DDPG control strategy formulation includes three parts: state space, action space, and reward function formulation, which are defined as follows: State Space: In formula (9), n ref , n are the reference speed and actual speed of the motor, rpm; i d_err , i q_err The motor d-axis reference current, the difference between the d-axis reference current and the actual current, the q-axis reference current, the difference between the d-axis reference current and the actual current, A; e ff The current efficiency of the electric drive system, 0-1 Action Space: A=[Δu d ,D d ] (10) In formula (10), Δu d ,Δu d is the motor d and q axis voltage correction value, -40-40; The reward function is: In formula (11), α, β, γ, χ, ζ are weights; i d_err 、i q_err The difference between the motor d-axis reference current and the actual current, the difference between the d-axis reference current and the actual current, A; e a 、e b is the efficiency value of the electric drive system before and after the motor, 0-1; Δu d ,Δu d is the motor d and q axis voltage correction value, -40-40.

10. The method for optimizing the efficiency of a permanent magnet synchronous motor based on optimal vehicle efficiency according to claim 9, characterized in that: Step S32, training and simulation analysis of the decision method: The CLTC condition is used as the training condition for the FW-DDPG control strategy: In order to fully train the FW-DDPG control strategy and shorten the algorithm training time, the motor external characteristic is divided into several specific operating points for training within the motor external characteristic curve; After the training is completed, the MAP diagram of the permanent magnet synchronous motor under the FW-DDPG control strategy is obtained.

Images