A collaborative optimization design method for vehicle transmission system and drive motor parameters

Abstract

The present invention belongs to the technical field of vehicle optimization design methods, and specifically relates to a collaborative optimization design method for vehicle transmission system and drive motor parameters. Based on vehicle parameters and performance indicators, with vehicle power, economy and motor performance as optimization targets, a high-precision machine learning prediction model for the motor is constructed, and a rapid estimation mathematical model for the motor efficiency Map is further constructed. Based on the high-precision machine learning prediction model for the motor and the rapid estimation mathematical model for the motor efficiency Map, a collaborative design optimization model for the vehicle transmission system and motor structural parameters is constructed. The method is solved by a multi-objective marine predator optimization algorithm, and the optimization of the electric vehicle transmission system is integrated with the motor structure optimization in an integrated design. The motor structural parameters are optimized with the goal of improving the vehicle drive efficiency, thereby effectively improving the vehicle power and economy, and at the same time reducing the cogging torque of the permanent magnet synchronous motor to reduce the vibration noise of the motor.

Description

Technical Field

[0001] The present invention belongs to the technical field of vehicle optimization design methods, and specifically relates to a collaborative optimization design method for vehicle transmission system and drive motor parameters. Background Art

[0002] With the rapid development of the new energy vehicle industry, its market size continues to expand. With advantages such as low emissions and high energy efficiency, new energy vehicles are gradually becoming a key component of modern transportation and transitioning to a more sustainable transportation future.

[0003] In the system design of new energy vehicles, the selection of each component and its parameter design are directly related to the power performance and operating efficiency of the vehicle. By optimizing the design of these component parameters, the acceleration and driving capabilities of the vehicle can be enhanced while reducing power consumption and comprehensively improving vehicle performance. In the prior art, some optimization design methods optimize the transmission parameters of the transmission system to achieve improved power and economy, but ignore the important role of the motor in the performance of the whole vehicle. Other optimization design methods use motor parameters as optimization variables and comprehensively consider the relationship between motor performance and vehicle performance. For the matching of power and economy of the transmission system, the above methods often use transmission parameters and motor parameters as optimization variables alone, without taking into account the matching of motor parameter design and vehicle performance, resulting in the inability to achieve optimal matching between motor performance and vehicle performance. Summary of the invention

[0004] The technical problem to be solved by the present invention is that the existing new energy vehicle optimization design method optimizes the transmission system and the motor structure separately, which limits the improvement of the vehicle driving efficiency. Therefore, a collaborative optimization design method for the vehicle transmission system and drive motor parameters is provided.

[0005] The present invention discloses a collaborative optimization design method for vehicle transmission system and drive motor parameters, comprising the following steps:

[0006] Step S1: construct a high-precision machine learning prediction model for the motor, the input of the model is the motor structural parameters, and the output is the rated torque and cogging torque of the motor;

[0007] Step S2: Based on the high-precision machine learning prediction model of the motor, a rapid estimation mathematical model of the motor efficiency map is constructed with the motor structural parameters as input;

[0008] Step S3: Based on the high-precision machine learning prediction model of the motor and the rapid estimation mathematical model of the motor efficiency map, the motor structural parameters, overload factor, expanded constant power factor and main reducer transmission ratio of the motor are taken as optimization design variables, and the 0-100km / h acceleration time, 15km / maximum climbing grade, WLTC working condition energy consumption and motor cogging torque are taken as optimization targets to build a collaborative design optimization model of the automobile transmission system and motor structural parameters;

[0009] Step S4: Solve the motor structure parameter collaborative design optimization model through a multi-objective marine predator optimization algorithm.

[0010] Furthermore, the step S1 includes the following steps:

[0011] Step S1.1: Select multiple motor structure parameter combinations to experimentally obtain the rated torque and cogging torque data sets of the motor under different combinations;

[0012] Step S1.2: Based on the rated torque and cogging torque data sets, select a variety of different machine learning models for training;

[0013] Step S1.3: Evaluate the prediction accuracy of different machine learning models, and select the machine learning model with the highest accuracy as the prediction model for the motor rated torque and cogging torque.

[0014] Furthermore, the step S2 includes the following steps:

[0015] Step S2.1: Based on the motor design parameters, a functional relationship between the motor design parameters and the motor loss is established;

[0016] Step S2.2: Based on the functional relationship between motor loss and motor efficiency, a motor efficiency map is established.

[0017] Furthermore, the functional relationship between the motor loss and the motor efficiency is:

[0018] ;

[0019] in, is the motor efficiency, , , , , and They are the motor’s mechanical power, mechanical loss, copper loss, hysteresis loss, eddy current loss and stray loss;

[0020] The motor efficiency map is represented as follows:

[0021] ;

[0022] in, is the rated speed, is the peak speed, is the rated torque, is the peak torque, is the peak efficiency.

[0023] Furthermore, the step S3 includes:

[0024] The slot width, axial length, permanent magnet width, permanent magnet length, pole arc coefficient, overload coefficient, expanded constant power coefficient and main reducer transmission ratio of the motor are taken as optimization design variables; 0-100km / h acceleration time, 15km / maximum climbing grade, WLTC condition energy consumption and motor cogging torque are taken as optimization targets; the maximum vehicle speed condition and ground adhesion condition are taken as constraints, and a collaborative multi-objective optimization model of vehicle transmission system-motor structure parameters is established.

[0025] Further, the step S4 includes:

[0026] Step S4.1: Calculate the fitness value of each solution of the vehicle transmission system-motor structure parameter collaborative multi-objective optimization model;

[0027] Step S4.2: traverse each solution and construct a set of non-dominated solutions, which is defined as a non-dominated multi-target prey cluster;

[0028] Step S4.3: Calculate the dominated number of each solution, and classify the solutions with a dominated number of zero into corresponding categories; repeatedly calculate the dominated numbers of the remaining solutions and classify them until all solutions are classified;

[0029] Step S4.4: construct a reward and penalty function;

[0030] Step S4.5: Each category randomly selects an iteration strategy;

[0031] Step S4.6: The solutions in the category are iterated according to the selected iteration strategy, the fitness of the solutions after iteration is calculated, and rewards or penalties are implemented according to the reward and penalty function, and the reward and penalty values ​​of the category are accumulated;

[0032] Step S4.7: Repeat step S4.3 to update the category of the solution after the division iteration;

[0033] Step S4.8: Select the iterative strategy with the highest reward-penalty value and repeat steps S4.6 and S4.7.

[0034] Step S4.9: Return to the non-multi-target prey cluster set and obtain the optimized parameters of the vehicle transmission system and drive motor.

[0035] Furthermore, the iteration strategy is an adaptive marine predator algorithm iteration strategy.

[0036] Beneficial effect: Based on the parameters and performance indicators of the whole vehicle, the present invention takes the power, economy and motor performance of the whole vehicle as the optimization goals, constructs a high-precision machine learning prediction model for the motor, further constructs a rapid estimation mathematical model for the motor efficiency Map, and constructs a collaborative design optimization model for the vehicle transmission system and the motor structure parameters based on the high-precision machine learning prediction model for the motor and the rapid estimation mathematical model for the motor efficiency Map. The model is solved by a multi-objective marine predator optimization algorithm, and the motor structure parameters are optimized with the goal of improving the whole vehicle driving efficiency by integrating the optimization of the electric vehicle transmission system with the optimization of the motor structure into an integrated design, thereby effectively improving the vehicle's power and economy, and at the same time reducing the cogging torque of the permanent magnet synchronous motor to reduce the vibration and noise of the motor. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0038] Figure 1 It is a schematic block diagram of the method flow of the present invention;

[0039] Figure 2 It is a schematic diagram of the reward and penalty function of the present invention;

[0040] Figure 3 This is a schematic diagram of reward and punishment areas in the iterative stage 1 of the present invention;

[0041] Figure 4 This is a schematic diagram of reward and punishment areas in the second iteration phase of the present invention;

[0042] Figure 5 This is a schematic diagram of the third reward and punishment area in the iterative stage of the present invention;

[0043] Figure 6 It is an iterative flow chart of the multi-objective marine predator algorithm of the present invention. DETAILED DESCRIPTION

[0044] In order to make the above-mentioned purposes, features and advantages of the present application more obvious and easy to understand, the specific implementation methods of the present application are described in detail below in conjunction with the accompanying drawings. In the following description, many specific details are set forth to facilitate a full understanding of the present application. However, the present application can be implemented in many other ways different from those described herein, and those skilled in the art can make similar improvements without violating the connotation of the present application, so the present application is not limited by the specific embodiments disclosed below.

[0045] This embodiment discloses a collaborative optimization design method for vehicle transmission system and drive motor parameters, comprising the following steps:

[0046] Step S1: construct a high-precision machine learning prediction model for the motor, the input of the model is the motor structural parameters, and the output is the rated torque and cogging torque of the motor;

[0047] Step S2: Based on the high-precision machine learning prediction model of the motor, a rapid estimation mathematical model of the motor efficiency map is constructed with the motor structural parameters as input;

[0048] Step S3: Based on the high-precision machine learning prediction model of the motor and the rapid estimation mathematical model of the motor efficiency map, the motor structural parameters, overload factor, expanded constant power factor and main reducer transmission ratio of the motor are taken as optimization design variables, and the 0-100km / h acceleration time, 15km / maximum climbing grade, WLTC working condition energy consumption and motor cogging torque are taken as optimization targets to build a collaborative design optimization model of the automobile transmission system and motor structural parameters;

[0049] Step S4: Solve the motor structure parameter collaborative design optimization model through a multi-objective marine predator optimization algorithm.

[0050] The present invention is based on the parameters and performance indicators of the whole vehicle, takes the power, economy and motor performance of the whole vehicle as the optimization goals, constructs a high-precision machine learning prediction model for the motor, further constructs a rapid estimation mathematical model for the motor efficiency Map, and constructs a collaborative design optimization model for the vehicle transmission system and the motor structure parameters based on the high-precision machine learning prediction model for the motor and the rapid estimation mathematical model for the motor efficiency Map. The model is solved by a multi-objective marine predator optimization algorithm, and the motor structure parameters are optimized with the goal of improving the whole vehicle driving efficiency by integrating the optimization of the electric vehicle transmission system with the optimization of the motor structure into an integrated design, thereby effectively improving the vehicle's power and economy, while reducing the cogging torque of the permanent magnet synchronous motor to reduce the vibration and noise of the motor.

[0051] Specifically, the step S1 includes the following steps:

[0052] Step S1.1: Select multiple motor structure parameter combinations and experimentally obtain the rated torque of the motor under different combinations ( ) and cogging torque ( ) dataset;

[0053] In the present invention, the motor structural parameters include the stator slot width , Motor axial length , Permanent magnet width , permanent magnet length and polar arc coefficient Each motor parameter structure includes three levels of preset values, so the full parameter sweep experiment is performed for a total of 3 5 Experiment.

[0054] Step S1.2: Based on the rated torque and cogging torque data sets, select a variety of different machine learning models for training;

[0055] In this embodiment, the machine learning models include random forest (RF), gradient boosting regression (GBR), K-nearest neighbor (KNN), linear regression (LR), Lasso regression (LASSO), adaptive boosting (AdaBoost), error back propagation neural network (BPNN), ridge regression (RR), radial basis support vector regression (SVR-RBF), Gaussian process regression (GPR), and Bayesian optimization support vector regression (BO-SVR).

[0056] Step S1.3: Evaluate the prediction accuracy of different machine learning models, and select the machine learning model with the highest accuracy as the prediction model for the motor rated torque and cogging torque.

[0057] In the present invention, the correlation coefficient is selected R 2 And mean absolute error MAE as the regression algorithm evaluation index, correlation coefficient R 2 Indicates the degree of fit between the prediction model and the true value. R 2 The closer the value is to 1, the better the model fit is. MAE represents the absolute difference between the predicted value and the true value. The smaller the MAE value is, the higher the prediction accuracy of the model is:

[0058] ;

[0059] ;

[0060] in, is the sample size, is the true value, is the model prediction value, is the average value.

[0061] The obtained high-precision machine learning prediction model of the motor is:

[0062] ;

[0063] ;

[0064] Specifically, step S2 includes the following steps:

[0065] Step S2.1: Based on the motor design parameters, a functional relationship between the motor design parameters and the motor loss is established;

[0066] The mechanical power of the motor at any operating point is:

[0067] ;

[0068] in, is the speed of the motor at this operating point, is the torque at this operating point.

[0069] The empirical formula for the mechanical loss and stray loss of the motor at any working point is as follows:

[0070] ;

[0071] ;

[0072] in, is the rated power of the motor.

[0073] At motor speed When , the hysteresis loss and eddy current loss are as follows:

[0074] ;

[0075] ;

[0076] in, is the rated speed of the motor, Peak efficiency point mechanical losses.

[0077] is the copper loss of the motor at the peak efficiency point, The calculation formula is as follows:

[0078] ;

[0079] At motor speed When , the formulas for hysteresis loss and eddy current loss are as follows:

[0080] ;

[0081] ;

[0082] At the rated operating point The copper loss is:

[0083] ;

[0084] The calculation of motor copper loss is divided into four cases:

[0085] when , torque When , the calculation formula is as follows:

[0086] ;

[0087] when , torque hour:

[0088] ;

[0089] when , torque hour:

[0090] ;

[0091] when , torque hour:

[0092] ;

[0093] in, is the rated torque, , is the correction factor, is the motor salient pole ratio, is the current angle of the motor in the constant power region, is the magnetic weakening coefficient, The motor is at rated power and at a speed of The torque when .

[0094] Motor constant power area current angle The calculation formula is as follows:

[0095] ;

[0096] Weak magnetic coefficient The calculation formula is as follows:

[0097] ;

[0098] in,

[0099] ;

[0100] ;

[0101] Step S2.2: Based on the functional relationship between motor loss and motor efficiency, a motor efficiency map is established.

[0102] The functional relationship between the motor loss and the motor efficiency is:

[0103] ;

[0104] in, is the motor efficiency, , , , , and They are the motor’s mechanical power, mechanical loss, copper loss, hysteresis loss, eddy current loss and stray loss;

[0105] Therefore, the motor efficiency map can be expressed as:

[0106] ;

[0107] in, is the rated speed, is the peak speed, is the rated torque, is the peak torque, is the peak efficiency.

[0108] The step S3 includes: taking the slot width, axial length, permanent magnet width, permanent magnet length, pole arc coefficient, overload coefficient, expanded constant power coefficient and main reducer transmission ratio of the motor as optimization design variables; taking 0-100km / h acceleration time, 15km / maximum climbing grade, WLTC condition energy consumption and motor cogging torque as optimization targets; taking the maximum vehicle speed condition and ground adhesion condition as constraints, and establishing a vehicle transmission system-motor structure parameter collaborative multi-objective optimization model.

[0109] Specifically, there is the following functional relationship between the motor rated power and peak power:

[0110] ;

[0111] in, is the rated power, is the motor overload factor;

[0112] The peak speed and rated speed have the following relationship:

[0113] ;

[0114] In the above formula Expand the constant power factor for the motor.

[0115] The slot width, axial length, permanent magnet width, permanent magnet length, pole arc coefficient, overload coefficient, expanded constant power coefficient and main reducer transmission ratio of the motor are used as optimization design variables, namely:

[0116] ;

[0117] 0-100km / h acceleration time , 15km / h maximum climbing grade 、WLTC condition energy consumption E and the motor cogging torque As optimization goal:

[0118] 0-100km / h acceleration time It means that the acceleration ability of pure electric vehicles on horizontal roads is mainly affected by driving force and driving resistance. The calculation formula is as follows:

[0119] ;

[0120] in, As driving force, is the air resistance, is the rolling resistance. Since the motor has two regions of constant torque and constant power, the driving force on the electric vehicle wheel is:

[0121] ;

[0122] in, is the peak torque of the motor, is the transmission efficiency, is the tire radius, is the rated speed of the motor, is the current speed of the motor.

[0123] ;

[0124] 15km / h maximum climbing grade , the maximum climbing grade of a pure electric vehicle is determined by the peak torque of the motor:

[0125] ;

[0126] WLTC condition energy consumption E， WLTC operating condition energy consumption refers to the energy consumed by an electric vehicle when driving a complete WLTC operating condition:

[0127] ;

[0128] in, is the vehicle speed at time t, is the efficiency of the motor at time t.

[0129] Through the WLTC operating diagram, according to the efficiency of the motor at any point The calculation formula can be used to obtain the motor efficiency of electric vehicles at any time point under WLTC conditions.

[0130] Motor cogging torque ,Based on the motor prediction model, the functional relationship between the motor cogging torque and the motor structural parameters can be known.

[0131] ;

[0132] In order to ensure that the optimized parameters can meet the requirements of electric vehicle power and economy, some constraints need to be set for the model:

[0133] Maximum speed constraint, the maximum speed should meet the maximum target speed designed in this paper ,Right now:

[0134] ;

[0135] Ground adhesion conditions: When the car is driving, it should meet the adhesion conditions, namely:

[0136] ;

[0137] in, is the pressure of the vehicle on the ground, is the ground adhesion coefficient, is the tire radius, For the transmission efficiency.

[0138] At this point, a collaborative multi-objective optimization model of vehicle transmission system-motor structural parameters is established.

[0139] In this embodiment, the marine predator algorithm sets prey and predator. The prey represents the individuals of the population, and these individuals move in the solution space through different strategies to find the best solution. The behavior of the predator imitates the process of predators looking for food in the real world, including random walks (exploration phase) and tracking prey (exploitation phase). The main task of the predator is to search the solution space and optimize the quality of the current solution through predation strategies. The entire prey population is recorded as ,in Represents the population size, Represents the dimension of the search space. The predator represents the best individual generated in each iteration. The predator concentrates on searching these areas by tracking the prey, thereby increasing the probability of finding the global optimal solution. During the optimization process, the position and characteristics of the prey may change continuously, reflecting the dynamic characteristics of the high-quality solution in the solution space. Its group is denoted by .

[0140] Specifically, the step S4 includes:

[0141] Step S4.1: Calculate the fitness value of each solution of the vehicle transmission system-motor structure parameter collaborative multi-objective optimization model;

[0142] In this embodiment, the solution of the initial population is the initial value of the given variable, and each solution is a variable vector x , distributed within the range of variables;

[0143] Step S4.2: traverse each solution and construct a set of non-dominated solutions, which is defined as a non-dominated multi-target prey cluster (NMPG);

[0144] Step S4.3: Calculate the dominated number of each solution, and classify the solutions with a dominated number of zero into corresponding categories; repeatedly calculate the dominated numbers of the remaining solutions and classify them until all solutions are classified;

[0145] In this embodiment, for each solution marked as the first frontier, all solutions dominated by it are found and the solutions of these solutions are A decrease of 1 in the value means that the domination number of these solutions decreases;

[0146] Continue to search from the remaining solutions = 0, label these solutions as the second category , repeat step S4.3. Each time a new frontier is marked, the remaining solutions The value continues to decrease, and the process is repeated until all solutions are assigned to a front;

[0147] Step S4.4: construct a reward and penalty function;

[0148] ;

[0149] Specifically, the schematic diagram of the reward-penalty function is as follows: Figure 2 As shown, The axes represent different areas, with the reward area located at , the penalty area is located in . The axes represent the reward and penalty values, with positive values ​​representing rewards and negative values ​​representing penalties. There are three regions for both the reward area and the penalty area, with each population corresponding to a separate area for each iteration, and rewards or penalties are given based on the iteration results.

[0150] Step S4.5: Each category randomly selects an iteration strategy;

[0151] In this embodiment, the iterative strategy in the adaptive marine predator algorithm is adopted, and each strategy is marked as arrive , as shown in the following table:

[0152] ;

[0153] in, Representative The step length of prey position update, represents Brownian motion, Flying on behalf of Levi, On behalf of the Levi-Brown Mixed Movement, represents the predator matrix, represents the prey matrix, symbol represents vector product, is a tuning parameter used to control the step size, Is a random operator that randomly generates a number between (0,1). It is an adaptive correction coefficient to control the position update step size. The coefficient decreases gradually as the iteration proceeds to ensure the convergence of the algorithm.

[0154] Step S4.6: The solutions in the category are iterated according to the selected iteration strategy, the fitness of the solutions after iteration is calculated, and rewards or penalties are implemented according to the reward and penalty function, and the reward and penalty values ​​of the category are accumulated;

[0155] In this embodiment, the population is given the ability to autonomously select iterative strategies through a reinforcement learning method, thereby realizing self-iteration of the multi-objective marine predator algorithm. First, the reward and penalty values ​​corresponding to the population completing the corresponding optimization strategy are defined, that is, when the population obtains better results after one iteration, a reasonable reward is given through the reward-penalty function, and vice versa, when a worse result is obtained, an appropriate penalty is given through the reward-penalty function. Each population will be given rewards and penalties for the specified area when selecting the next iterative strategy.

[0156] According to the reward and penalty function, the function randomly generates reward and penalty values ​​in the corresponding area. The reward and penalty intervals corresponding to each iteration stage are defined as follows: Figures 3-5 As shown,

[0157] symbol Defined as belonging to the NMPG category at the tth iteration and has an iterative strategy The reward and penalty value of Defined as Category Cumulative reward penalty value. Using the greedy strategy, each individual in the population selects the one with the maximum At the beginning of an iteration, Initialized to 0, each population randomly selects a strategy. Then, the various populations are classified according to their NMPG categories. Rewards and punishments are given according to the ranking. Repeating the above rules, each population can independently select the best iteration strategy to complete the optimization. This adaptive iteration process is as follows Figure 6 shown.

[0158] Step S4.7: Repeat step S4.3 to update the category of the solution after the division iteration;

[0159] Step S4.8: Select the iterative strategy with the highest reward-penalty value and repeat steps S4.6 and S4.7.

[0160] Step S4.9: Return to the non-multi-target prey cluster set and obtain the optimized parameters of the vehicle transmission system and drive motor.

[0161] In this embodiment, the constructed multi-objective marine predator algorithm has three stages in its iteration process, each of which involves a different optimization strategy. During the iteration process, the position of each population is recorded and evaluated, and rewards and punishments are given accordingly according to the reward and punishment strategy to achieve self-iteration. As the iteration proceeds, the search space gradually shrinks and converges.

[0162] In this embodiment, the iteration strategy is an adaptive marine predator algorithm iteration strategy, the fish school effect coefficient of the multi-target marine predator algorithm is 0.2, and the correction coefficient is 0.5.

[0163] The technical features of the above-described embodiments may be arbitrarily combined. To make the description concise, not all possible combinations of the technical features in the above-described embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0164] The above-described embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be construed as limiting the scope of the patent application. It should be pointed out that, for a person of ordinary skill in the art, several variations and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the patent application shall be subject to the attached claims.

Claims

1. A collaborative optimization design method for vehicle transmission system and drive motor parameters, characterized in that: The following steps are involved: Step S1: construct a high-precision machine learning prediction model for the motor, the input of the model is the motor structural parameters, and the output is the rated torque and cogging torque of the motor; Step S2: Based on the high-precision machine learning prediction model of the motor, a rapid estimation mathematical model of the motor efficiency map is constructed with the motor structural parameters as input; Step S3: Based on the high-precision machine learning prediction model of the motor and the rapid estimation mathematical model of the motor efficiency map, the motor structural parameters, overload factor, expanded constant power factor and main reducer transmission ratio of the motor are taken as optimization design variables, and the 0-100km / h acceleration time, 15km / maximum climbing grade, WLTC working condition energy consumption and motor cogging torque are taken as optimization targets to build a collaborative design optimization model of the automobile transmission system and motor structural parameters; Step S4: solving the motor structure parameter collaborative design optimization model by using a multi-objective marine predator optimization algorithm; Step S4.1: Calculate the fitness value of each solution of the vehicle transmission system-motor structure parameter collaborative multi-objective optimization model; Step S4.2: traverse each solution and construct a set of non-dominated solutions, which is defined as a non-dominated multi-target prey cluster; Step S4.3: Calculate the dominated number of each solution, and divide the solutions with a dominated number of zero into corresponding categories; Repeatedly calculate the dominated numbers of the remaining solutions and classify them until all solutions are classified; Step S4.4: construct a reward and penalty function; Step S4.5: Each category randomly selects an iteration strategy; Step S4.6: The solutions in the category are iterated according to the selected iteration strategy, the fitness of the solutions after iteration is calculated, and rewards or penalties are implemented according to the reward and penalty function, and the reward and penalty values ​​of the category are accumulated; Step S4.7: Repeat step S4.3 to update the category of the solution after the division iteration; Step S4.8: Select the iteration strategy with the highest reward-penalty value and repeat steps S4.6 and S4.7; Step S4.9: Return to the non-multi-target prey cluster set and obtain the optimized parameters of the vehicle transmission system and drive motor.

2. The method for collaborative optimization design of vehicle transmission system and drive motor parameters according to claim 1, characterized in that: The step S1 comprises the following steps: Step S1.1: Select multiple motor structure parameter combinations to experimentally obtain the rated torque and cogging torque data sets of the motor under different combinations; Step S1.2: Based on the rated torque and cogging torque data sets, select a variety of different machine learning models for training; Step S1.3: Evaluate the prediction accuracy of different machine learning models, and select the machine learning model with the highest accuracy as the prediction model for the motor rated torque and cogging torque.

3. The method for collaborative optimization design of vehicle transmission system and drive motor parameters according to claim 1, characterized in that: The step S2 comprises the following steps: Step S2.1: Based on the motor design parameters, a functional relationship between the motor design parameters and the motor loss is established; Step S2.2: Based on the functional relationship between motor loss and motor efficiency, a motor efficiency map is established.

4. The method for collaborative optimization design of vehicle transmission system and drive motor parameters according to claim 3, characterized in that: The functional relationship between the motor loss and the motor efficiency is: ； in, is the motor efficiency, , , , , and They are the motor’s mechanical power, mechanical loss, copper loss, hysteresis loss, eddy current loss and stray loss; The motor efficiency map is represented as follows: ； in, is the rated speed, is the peak speed, is the rated torque, is the peak torque, is the peak efficiency.

5. The method for collaborative optimization design of vehicle transmission system and drive motor parameters according to claim 1, characterized in that: The step S3 comprises: The slot width, axial length, permanent magnet width, permanent magnet length, pole arc coefficient, overload coefficient, expanded constant power coefficient and main reducer transmission ratio of the motor are taken as optimization design variables; 0-100km / h acceleration time, 15km / maximum climbing grade, WLTC condition energy consumption and motor cogging torque are taken as optimization targets; the maximum vehicle speed condition and ground adhesion condition are taken as constraints, and a collaborative multi-objective optimization model of vehicle transmission system-motor structure parameters is established.

6. The method for collaborative optimization design of vehicle transmission system and drive motor parameters according to claim 1, characterized in that: The iteration strategy is an adaptive marine predator algorithm iteration strategy.

Images