Heat dissipation method and system for vehicle electric drive assembly, electronic equipment and storage medium

By building and optimizing the proxy model of the electric drive assembly and using the big data multi-objective optimization algorithm, the problem of poor heat dissipation effect of the electric drive assembly in the existing technology is solved, and high-precision heat dissipation effect and extended motor life are achieved.

CN118940635AActive Publication Date: 2024-11-12CHONGQING JINKANG POWER NEW ENERGY CO LTD
View PDF 3 Cites 0 Cited by

Patent Information

Application Number
CN202411063876.0
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-08-05
Publication Date
2024-11-12
Estimated Expiration
2044-08-05

AI Technical Summary

Technical Problem

The prior art rarely optimizes the cooling system of the vehicle electric drive assembly through big data multi-objective optimization, resulting in poor heat dissipation effect.

Method used

By obtaining sample space with reasonable distribution of sample points, an electric drive assembly agent model is constructed, and the maximum temperature and coolant pressure of the electric drive assembly are optimized to improve the heat dissipation effect.

Benefits of technology

A high-precision heat dissipation model is realized, and the R-square value reaches 0.983, which significantly improves the heat dissipation effect of the electric drive assembly, provides theoretical support for the actual electric drive assembly heat dissipation system, and extends the service life of the motor.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118940635A_ABST
    Figure CN118940635A_ABST
Patent Text Reader

Abstract

The invention discloses a heat dissipation method and system for a vehicle electric drive assembly, electronic equipment and a storage medium. The heat dissipation method for the vehicle electric drive assembly comprises the steps that a sample space with reasonably distributed sample points is obtained; according to the sample space with the reasonably distributed sample points, selecting the highest temperature, the cooling water pressure and the cooling oil pressure of the electric drive assembly as research targets, and constructing an electric drive assembly agent model; the electric drive assembly agent model is optimized through an incremental learning strategy, and a high-precision agent model is obtained; and solving the high-precision proxy model by using a multi-objective optimization algorithm by taking an oil path structure of a motor as a design variable and taking the highest temperature of an electric drive assembly and the pressure of cooling liquid as optimization objectives to obtain optimal solutions of different optimization algorithms. The method is verified to have high reliability, the heat dissipation effect of the electric drive assembly can be remarkably improved, theoretical support is provided for production and processing of an actual heat dissipation system of the electric drive assembly, internal heat dissipation of the motor can be improved, and the service life of the motor is prolonged.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of vehicle technology, and more specifically, to a heat dissipation method, system, electronic device and storage medium for a vehicle electric drive assembly. Background Art

[0002] As an environmentally friendly secondary energy source, electricity can replace fossil fuels in traditional internal combustion engines, drive motors to output mechanical energy, and vigorously develop new energy vehicles. It is an important way to solve energy and environmental problems, and is also a key issue in the national development plan.

[0003] The electric drive assembly driven by the power system has the characteristics of high power density, high efficiency, fast response speed and wide speed regulation range, and has gradually become the competitive core of the electric drive assembly driven by new energy vehicles. Optimizing the design of the heat dissipation system for the developed high-power electric drive assembly products, finding the optimal heat dissipation parameters of the electric drive assembly, and obtaining the best heat dissipation effect are one of the biggest difficulties in the development process. Existing design technologies often optimize the structure by drawing on mature products on the market or a large amount of performance development test data, and rarely establish development models through big data multi-objective optimization. Therefore, it is a technical problem that needs to be solved in this field. Summary of the invention

[0004] In view of this, the present invention provides a heat dissipation method for a vehicle electric drive assembly, so as to solve the problem that the prior art rarely establishes a development model through a big data multi-objective optimization method.

[0005] In a first aspect, the present application provides a method for heat dissipation of a vehicle electric drive assembly, comprising:

[0006] Obtain a sample space with reasonable distribution of sample points;

[0007] According to the sample space with reasonable distribution of sample points, the maximum temperature, cooling water pressure and cooling oil pressure of the electric drive assembly are selected as research targets, and the proxy model of the electric drive assembly is constructed;

[0008] The electric drive assembly proxy model is optimized through an incremental learning strategy to obtain a high-precision proxy model.

[0009] Taking the oil circuit structure of the motor as the design variable and the maximum temperature and coolant pressure of the electric drive assembly as the optimization targets, the multi-objective optimization algorithm is used to solve the high-precision proxy model to obtain the optimal solutions of different optimization algorithms.

[0010] In a second aspect, the present application provides a heat dissipation system for a vehicle electric drive assembly, which is applied to the heat dissipation method for the above-mentioned vehicle electric drive assembly, including:

[0011] An acquisition module is used to obtain a sample space with a reasonable distribution of sample points;

[0012] A construction module is used to select the maximum temperature, cooling water pressure and cooling oil pressure of the electric drive assembly as research targets according to the sample space with reasonable sample point distribution, and to construct an electric drive assembly proxy model;

[0013] An optimization module is used to optimize the electric drive assembly proxy model through an incremental learning strategy to obtain a high-precision proxy model;

[0014] The calculation module is used to solve the high-precision proxy model using the oil circuit structure of the motor as the design variable and the maximum temperature and coolant pressure of the electric drive assembly as the optimization targets using a multi-objective optimization algorithm to obtain the optimal solution of different optimization algorithms.

[0015] In a third aspect, the present application provides an electronic device, the electronic device comprising:

[0016] memory for storing programs; and

[0017] A processor is used to execute any of the above-mentioned methods for heat dissipation of a vehicle electric drive assembly by calling a program stored in a memory.

[0018] In a fourth aspect, the present application provides a computer-readable storage medium, characterized in that the computer-readable storage medium stores computer instructions, and the computer instructions are used to enable a computer to execute any of the above-mentioned methods for dissipating heat of a vehicle electric drive assembly.

[0019] Compared with the prior art, the heat dissipation method, system, electronic device and storage medium of the vehicle electric drive assembly provided by the present invention achieve at least the following beneficial effects:

[0020] The present invention provides a heat dissipation method, system, electronic device and storage medium for a vehicle electric drive assembly. The heat dissipation method for the vehicle electric drive assembly includes: obtaining a sample space with a reasonable distribution of sample points; according to the sample space with a reasonable distribution of sample points, selecting the maximum temperature, cooling water pressure and cooling oil pressure of the electric drive assembly as research targets, and constructing an electric drive assembly proxy model; optimizing the electric drive assembly proxy model through an incremental learning strategy to obtain a high-precision proxy model; using the oil circuit structure of the motor as a design variable, and the maximum temperature and coolant pressure of the electric drive assembly as optimization targets, solving the high-precision proxy model using a multi-objective optimization algorithm, and obtaining the optimal solutions of different optimization algorithms. Using the above scheme, the electric drive assembly proxy model is optimized through an incremental learning strategy. The optimization was performed, which significantly improved the accuracy of the high-precision proxy model. The R-square value in the final evaluation index could reach 0.983, which can truly reflect the actual physical situation. Then, different multi-objective optimization algorithms were used to solve the high-precision proxy model, and the Pareto optimal solutions of different multi-objective optimization algorithms were obtained. The optimal solutions of different multi-objective optimization algorithms were obtained through the Utopia point method. It was verified that the optimal solution has high reliability and can significantly improve the heat dissipation effect of the electric drive assembly, providing theoretical support for the production and processing of the actual electric drive assembly heat dissipation system. In addition, since the service life of the induction motor is greatly affected by the maximum operating temperature of the electric drive assembly, improving the internal heat dissipation of the motor can increase the life of the motor.

[0021] Of course, any product implementing the present invention does not necessarily need to achieve all of the above technical effects at the same time.

[0022] Further features and advantages of the present invention will become apparent from the following detailed description of exemplary embodiments of the present invention with reference to the attached drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] The accompanying drawings, which are incorporated in and constitute a part of the specification, illustrate embodiments of the invention and, together with the description, serve to explain the principles of the invention.

[0024] Figure 1 It is a schematic flow chart of a heat dissipation method for a vehicle electric drive assembly provided by the present invention;

[0025] Figure 2 It is a structural schematic diagram of a heat dissipation device for a vehicle electric drive assembly provided by the present invention;

[0026] Figure 3 It is a cross-sectional schematic diagram of a heat dissipation device for a vehicle electric drive assembly provided by the present invention;

[0027] Figure 4 It is a sample point distribution diagram of a BBD provided by the present invention;

[0028] Figure 5It is a sample point distribution diagram of a CCD provided by the present invention;

[0029] Figure 6 It is a sample point distribution diagram of LHS provided by the present invention;

[0030] Figure 7 It is a sample point distribution diagram of an OFS provided by the present invention;

[0031] Figure 8 It is a real sampling sample point distribution map provided by the present invention;

[0032] Fig. 9 It is a bar chart of the accuracy of heat dissipation parameters (cooling oil pressure) of the electric drive assembly under different proxy models provided by the present invention;

[0033] Fig.10 It is a bar chart of the accuracy of heat dissipation parameters (cooling water pressure) of the electric drive assembly under different proxy models provided by the present invention;

[0034] Fig.11 It is a bar chart of the accuracy of the heat dissipation parameters (maximum temperature of the electric drive assembly) of the electric drive assembly under different proxy models provided by the present invention;

[0035] Fig.12 It is a flow chart of a high-precision proxy model provided by the present invention;

[0036] Fig.13 It is a high-precision proxy model improvement accuracy map provided by the present invention;

[0037] Fig.14 is an optimal solution set of a different multi-objective algorithm provided by the present invention;

[0038] Fig.15 It is a temperature isothermal curve cloud diagram of an electric drive assembly provided by the present invention (side view of the initial scheme);

[0039] Fig.16 It is a temperature isothermal curve cloud diagram of an electric drive assembly provided by the present invention (front view of the initial scheme);

[0040] Fig.17 It is a temperature isothermal curve cloud diagram of an electric drive assembly provided by the present invention (a side view of the optimal solution);

[0041] Fig.18 It is a temperature isothermal curve cloud diagram of an electric drive assembly provided by the present invention (a side view of the optimal solution);

[0042] Fig.19 It is a temperature isothermal curve cloud diagram (temperature label) of an electric drive assembly provided by the present invention;

[0043] Fig. 20It is a photo of the experimental process in the experimental verification provided by the present invention;

[0044] Fig.21 It is the experimental result curve in the experimental verification provided by the present invention;

[0045] Fig. 22 It is a structural schematic diagram of a heat dissipation optimization framework for an electric drive assembly provided by the present invention;

[0046] Fig.23 It is a structural schematic diagram of a heat dissipation system of a vehicle electric drive assembly provided by the present invention. DETAILED DESCRIPTION

[0047] Various exemplary embodiments of the present invention will now be described in detail with reference to the accompanying drawings. It should be noted that the relative arrangement of components and steps, numerical expressions and numerical values ​​set forth in these embodiments do not limit the scope of the present invention unless otherwise specifically stated.

[0048] The following description of at least one exemplary embodiment is merely illustrative in nature and is in no way intended to limit the invention, its application, or uses.

[0049] Technologies, methods, and equipment known to ordinary technicians in the relevant art may not be discussed in detail, but where appropriate, the technologies, methods, and equipment should be considered part of the specification.

[0050] In all examples shown and discussed herein, any specific values ​​should be interpreted as merely exemplary and not limiting. Therefore, other examples of the exemplary embodiments may have different values.

[0051] It should be noted that like reference numerals and letters refer to similar items in the following figures, and therefore, once an item is defined in one figure, it need not be further discussed in subsequent figures.

[0052] Embodiment 1

[0053] Reference Figure 1-Figure 21 As shown, Figure 1 It is a schematic flow chart of a heat dissipation method for a vehicle electric drive assembly provided by the present invention; Figure 2 It is a structural schematic diagram of a heat dissipation device for a vehicle electric drive assembly provided by the present invention; Figure 3 It is a cross-sectional schematic diagram of a heat dissipation device for a vehicle electric drive assembly provided by the present invention;

[0054] Figure 4 It is a sample point distribution diagram of a BBD provided by the present invention; Figure 5 It is a sample point distribution diagram of a CCD provided by the present invention; Figure 6 It is a sample point distribution diagram of LHS provided by the present invention; Figure 7 It is a sample point distribution diagram of an OFS provided by the present invention; Figure 8 It is a real sampling sample point distribution map provided by the present invention; Fig. 9 It is a bar chart of the accuracy of heat dissipation parameters (cooling oil pressure) of the electric drive assembly under different proxy models provided by the present invention; Fig.10 It is a bar chart of the accuracy of heat dissipation parameters (cooling water pressure) of the electric drive assembly under different proxy models provided by the present invention; Fig.11 It is a bar chart of the accuracy of the heat dissipation parameters (maximum temperature of the electric drive assembly) of the electric drive assembly under different proxy models provided by the present invention;

[0055] Fig.12 It is a flow chart of a high-precision proxy model provided by the present invention; Fig.13 It is a high-precision proxy model improvement accuracy map provided by the present invention; Fig.14 is an optimal solution set of a different multi-objective algorithm provided by the present invention; Fig.15 It is a temperature isothermal curve cloud diagram of an electric drive assembly provided by the present invention (side view of the initial scheme); Fig.16 It is a temperature isothermal curve cloud diagram of an electric drive assembly provided by the present invention (front view of the initial scheme); Fig.17 It is a temperature isothermal curve cloud diagram of an electric drive assembly provided by the present invention (a side view of the optimal solution); Fig.18 It is a temperature isothermal curve cloud diagram of an electric drive assembly provided by the present invention (a side view of the optimal solution); Fig.19 It is a temperature isothermal curve cloud diagram (temperature label) of an electric drive assembly provided by the present invention; Fig. 20 It is a photo of the experimental process in the experimental verification provided by the present invention; Fig.21 is an experimental result curve in the experimental verification provided by the present invention; this embodiment provides a heat dissipation method for a vehicle electric drive assembly, including:

[0056] S1 obtains a sample space with a reasonable distribution of sample points;

[0057] Firstly, the heat dissipation design variables corresponding to the electric drive assembly are obtained, and the heat dissipation design variables include oil inner diameter, nozzle size, water channel connection width, oil inlet diameter, oil inlet depth, cooling oil flow rate and cooling water flow rate; secondly, the heat dissipation design variables are sampled by using the experimental design method, and a sample space with a reasonable distribution of sample points is selected. Optionally, the sample space with a reasonable distribution of sample points can be a sample space with an optimal distribution of sample points;

[0058] Specifically, combined Figure 2 and Figure 3As shown in the figure, when constructing the proxy model of the electric drive assembly, it is necessary to first determine the key parameters that affect the heat dissipation research. Through the research on the parameters of the electric drive assembly and the heat dissipation system, it can be found that the liquid flow rate, water jacket structure and oil circuit structure have a great influence on the heat dissipation system of the electric drive assembly. Considering the actual heat dissipation research value and product processing capabilities, the oil inner diameter r, nozzle size c, water channel connection width b, oil inlet diameter R, oil inlet depth L, cooling oil flow rate V o and cooling water flow rate V w As design variables, the design variables [L, R, r, b, h, V o ,V w The upper and lower limits of ] are designed as [18, 6, 15, 24, 10, 0.5, 0.2] and [22, 9, 25, 28, 20, 1.5, 0.8] respectively, which are obtained based on the analysis of influencing factors and the corresponding actual product processing capabilities. In order to clearly express the influence of design variables on the heat dissipation of the electric drive assembly, the maximum temperature T of the electric drive assembly is selected. max , Cooling water pressure P w and cooling oil pressure P o The research goal is to dissipate heat for the electric drive assembly, which is a factor directly related to heat dissipation capacity and cost.

[0059] After determining the corresponding heat dissipation design variables, it is necessary to conduct experimental design and sample the design variables. Currently, the commonly used experimental design methods include response surface experimental design (Box-Behnken Design Sampling, BBD), central combination experimental design sampling (Central Combination Design Sampling, CCD), Latin Hypercube Experimental Design Sampling (Latin Hypercube Design Sampling, LHS) and optimal space filling experimental (Optimal Space-Filling Design Sampling, OSF) design methods. The introduction and characteristics of the corresponding methods are shown in Table 1.

[0060] Table 1 Experimental design method

[0061]

[0062] The samples of the above four sampling methods are as follows Figure 4-Figure 7As shown. Analysis shows that since the sampling points of BBD and CCD are automatically generated according to the space, their distribution effect is not as good as LHS and OSF, that is, the sampling points of BBD and CCD cannot effectively fill the sampling space. At the same time, the sampling points of LHS and OSF can effectively fill the sample space. In order to evaluate the adaptability of these two sampling methods to this problem, the coordinate exchange algorithm is introduced here. First, assume that the sampling result is:

[0063]

[0064] The definition of the evaluation index G is shown in the formula. The smaller the G value is, the better the uniformity and spatial distribution of the sampling results are.

[0065]

[0066] After calculation, we know that G(LHS) = 1542.64, G(OSF) = 1673.18. Therefore, it can be found that LHS is more adaptable to this problem, and the sample points are more evenly and fully distributed in the sample space.

[0067] The specific sampling process of sample point collection is as follows, and the final samples are shown in Table 2 and Figure 8 shown.

[0068] Step 1: Determine the required number of samples N = 100 and the sample size D = 7;

[0069] Step 2: Divide the [0,1] interval into 100 segments and randomly select a value in any segment of each dimension;

[0070] Step 3: Combine the 7-dimensional values ​​into a vector and map it to the standard sample value through the inverse function of the standard normal distribution;

[0071] Step 4: Repeat steps 2-3. During the repetition, for a single dimension, the segments that have been sampled will no longer be sampled. Sample all segments of all dimensions. At this point, sampling is complete and 100 sampling points are obtained.

[0072] Table 2 Data of some sample points and target values

[0073]

[0074]

[0075] It should be noted that in Table 2 above, r is the inner diameter of the oil; c is the nozzle size; b is the water channel connection width; R is the oil inlet hole diameter; L is the oil inlet depth; V o is the cooling oil flow rate; V wis the cooling water flow rate; T max is the maximum temperature of the electric drive assembly; P w is the cooling water pressure; P o To cool the oil pressure.

[0076] S2 selects the maximum temperature, cooling water pressure and cooling oil pressure of the electric drive assembly as the research targets according to the sample space with reasonable sample point distribution, and constructs the proxy model of the electric drive assembly;

[0077] Specifically, for complex engineering problems such as heat dissipation, it is often difficult to directly obtain the functional expression between input and output. In order to overcome this difficulty, data-driven surrogate model methods have been widely used in many fields. Based on the design of experiments, sample points obtained from the experimental design are used to establish surrogate models and fit the relationship between input and output. Common surrogate model methods include response surface analysis RSM, radial basis function algorithm RBF, support vector regression algorithm SVR and Kriging interpolation Kriging method.

[0078] RSM is a simple and effective surrogate model that fits the relationship between input and output through a polynomial. The most appropriate accuracy of the model can be adjusted according to the order of the polynomial. The basic expression is as follows:

[0079]

[0080] Among them, x i represents the design variables involved in fitting; β0, β i , β ij Represents the coefficients of the regression equation; y1, y2, y3, and y4 correspond to the design variable response values ​​of the first-order, second-order, third-order, and fourth-order response surface methods.

[0081] RBF is a single-layer neural network with strong generalization ability and excellent convergence speed. Its basic function expression is:

[0082]

[0083] Among them, ω i is the model adaptability weight coefficient, Ψ(·) is a nonlinear basis function representing the Euclidean distance between two sampling points, and y is the output value.

[0084] SVR realizes data regression prediction through hyperplane and support vector. Its core is to transform the problem into a quadratic programming problem. Its equation is:

[0085]

[0086] Among them, a iis the Lagrange multiplier, and Φ is a nonlinear transformation function that can map training samples to a high-dimensional space.

[0087] S3 optimizes the electric drive assembly proxy model through an incremental learning strategy to obtain a high-precision proxy model;

[0088] Incremental learning, also known as continuous learning or lifelong learning, is a machine learning method that allows the model to continuously learn from new data instead of retraining the entire model from scratch. This method allows the model to continuously learn new knowledge and adapt to changes in the ever-changing environment.

[0089] Firstly, the accuracy of the electric drive assembly proxy model is evaluated by using evaluation indicators. Secondly, the incremental Kriging method is used to optimize the design variables with insufficient accuracy in the evaluated electric drive assembly proxy model. The initial training set is continuously optimized and updated through the expected improvement method to obtain a high-precision proxy model.

[0090] Specifically, in order to verify the accuracy of the proxy model, evaluation indicators are introduced. Median absolute deviation (MAD), maximum absolute error (MAE), root mean square error (RMSE) and R2 are four typical evaluation indicators that can fully reflect the accuracy of the electric drive assembly proxy model fitting. Among them, the closer R-square is to 1, or the closer MAD, MAE, and RMSE are to 0, the higher the prediction accuracy of the electric drive assembly proxy model.

[0091]

[0092]

[0093]

[0094]

[0095] In the formula, y is the predicted value, is the true value, median is the median, N is the number of samples in the test set, and mean is the average value.

[0096] The first 80 samples are used as the training set and the last 20 samples are used as the test set. The prediction accuracy of different modeling methods is shown in Figure 9-11 shown.

[0097] Through the comparison results between RSM, RBF, SVR and Kriging, it can be found that the Kriging model has the highest accuracy for the three parameters (maximum temperature of the electric drive assembly, cooling oil pressure and cooling water pressure). For the research target of cooling water pressure, although the Kriging model has the best modeling accuracy among the existing methods, its accuracy is still relatively low. Therefore, corresponding methods are needed to improve the accuracy of the corresponding Kriging model.

[0098] In the problem of high-dimensional data fitting, the traditional proxy model method often fails to find the optimal fitting space due to the limitation of the initial experimental design, which leads to the low accuracy of the proxy model. In order to improve this phenomenon, a method called adaptive sequential sampling has been widely used. This is a strategy based on incremental learning. By continuously optimizing and updating the initial training set, the purpose of improving the model accuracy is finally achieved. In order to better improve the modeling accuracy of the research target of cooling water pressure, a Kriging proxy model method based on sequential sampling is used here, also known as the incremental Kriging method (Incremental Kriging, IK). Its core lies in continuously updating the corresponding process variables during the modeling process.

[0099] Assuming that the number of new sample points added is s, the incremental Kriging expression is:

[0100]

[0101] Among them, the relevant update process variables are as follows:

[0102]

[0103]

[0104] In the formula, is a matrix composed of the values ​​of the original sample points and the incremental sample points, which is used to solve the incremental kriging method above; y t T is the original sample point matrix, y n+s n+1 to n+s represent s newly added sample points.

[0105]

[0106]

[0107] In the formula, is a matrix composed of the covariance of the input variables, used to solve the Kriging method, and R is the covariance between the input design variables.

[0108] The expected improvement method is used to increase sample points in the incremental kriging method. Assume that for any design variable x, the response value y is a random variable that obeys a normal distribution and its mean is Its probability density function is:

[0109]

[0110] Assume that the optimal response value of the current sample is y b , then the expected value of the objective function at this design point is:

[0111]

[0112] The expected improved function is as follows:

[0113]

[0114] Where φ is the cumulative distribution function of the standard normal distribution; is the standard deviation, is the response value based on the design variable x.

[0115] By analyzing and deducing the expected improvement function, it can be found that the expected improvement function is about is a monotonically decreasing function, that is, when When the value of decreases, the value of I(x) will increase accordingly, so it is recommended to add sampling points in the minimum response value area of ​​the sample. At the same time, the expected improvement function is also y b A monotonically increasing function, when y b When increases, I(x) will have a larger expected value, that is, in a larger area, the sample points are relatively scarce, with less sample information and model uncertainty. Here, additional sampling is also needed to improve the model.

[0116] In summary, refer to Fig.12 As shown in the figure, the basic process of the IK model includes: first, determine the design variables and design goals, where the design variables include oil inner diameter r, nozzle size c, water channel connection width b, oil inlet diameter R, oil inlet depth L, cooling oil flow rate v o and cooling water flow rate v w ; Design objectives include the maximum temperature Tmax of the electric drive assembly, the cooling water pressure P w and cooling oil pressure P O ;

[0117] Then, based on the experimental design method, the design variables are sampled. Among them, the LHS experimental design sampling method in this embodiment has stronger adaptability and more uniform sample distribution, so the sampling results of this method are applicable;

[0118] Secondly, after sampling is completed, a series of sampling point data are obtained, and a proxy model method is used to fit the relationship between the input design variables of the sample points and the output design targets. In this embodiment, the initial Kriging model can be used as the proxy model;

[0119] Evaluation indicators such as median absolute deviation (MAD), maximum absolute error (MAE), root mean square error (RMSE) and R2 are selected to reflect the accuracy of model fitting. For areas where the Kriging model is not accurate enough (water pressure is used as an example in this embodiment), the IK method (incremental Kriging method) is used to improve the modeling, and the EI method (expected improvement method) is used to add new training samples. The accuracy of the JK model finally trained is as follows: Fig.13 As shown in the figure, it can be found that with the increase of sample points, the accuracy of the Kriging model has been significantly improved, and the final R-square value can reach 0.983, which can truly reflect the actual physical situation.

[0120] It should be noted that the R-square value in the evaluation index may vary according to different projects, and this embodiment does not make any specific limitation on this.

[0121] (1) Heat dissipation optimization model for electric drive assembly

[0122] ① Design variables

[0123] The design variables are mainly divided into the structural parameters of the heat dissipation system and the cooling medium parameters. Among them, the structural parameters mainly include five parameters: oil inlet depth L, oil inlet diameter R, oil inlet inner diameter r, nozzle size c, and water channel connection width b. The cooling medium parameters include the cooling water flow rate V w , cooling oil flow rate V o There are two types of parameters, which are related variable parameters designed based on the comparison of cooling solutions.

[0124] ② Research objectives

[0125] Because the purpose of this study is to achieve efficient heat dissipation of the electric drive assembly, the maximum temperature of the electric drive assembly T max , cooling water pressure P w and cooling oil pressure P o As a corresponding research target, the maximum temperature of the electric drive assembly T max The research goal directly reflects the heat dissipation effect of the electric drive assembly. The cooling water pressure P w and cooling oil pressure P o The research objectives are associated with the corresponding heat dissipation costs, which can comprehensively reflect the optimal heat dissipation of the electric drive assembly.

[0126] ③Electric drive assembly agent model

[0127] After analyzing the corresponding design variables and research objectives, the electric drive assembly agent model is shown as follows:

[0128]

[0129] S4 uses the oil circuit structure of the motor as the design variable, the maximum temperature and coolant pressure of the electric drive assembly as the optimization targets, and uses a multi-objective optimization algorithm to solve the high-precision proxy model to obtain the optimal solution of different optimization algorithms.

[0130] In order to ensure the reliability of the optimization results, the electric drive assembly proxy model is optimized by a multi-objective optimization algorithm, which includes the multi-objective genetic algorithm NSGA-II, the sparse large-scale multi-objective evolutionary algorithm SparseEA, TOP (used to solve complex problems and optimization algorithms) and the multi-objective particle swarm algorithm MOPSO. NSGA-II is a classic multi-objective genetic algorithm that quickly and accurately finds the optimal area through non-dominated sorting and elite strategies. SparseEA is an evolutionary algorithm for sparse data and has been widely used in large-scale sparse matrix optimization processes. TOP is a two-stage optimization method that transforms multi-objective problems into single-objective problems. MOPSO applies the particle swarm algorithm in the multi-objective field and achieves optimization effects by marking and updating particles.

[0131] The optimization results of the above four multi-objective optimization algorithms are as follows: Fig.14 As shown in Figure 2, the results show that the solution set of MOPSO is too dense and cannot well display the optimal space. Except for the MOPSO method, the other three methods can obtain good Pareto solutions.

[0132] Due to the three R&D goals (maximum temperature of the electric drive assembly T max , cooling water pressure P w and cooling oil pressure P o ), it is impossible to obtain three optimal solutions at the same time. Since the Pareto solution set generated by the multi-objective genetic algorithm has a non-dominated relationship, it is impossible to compare the advantages and disadvantages of different solutions. Therefore, the Utopia point method is introduced to obtain the actual solution required by the electric drive assembly.

[0133] Taking the oil circuit structure of the motor as the design variable, the maximum temperature and coolant pressure of the electric drive assembly as the optimization target, the high-precision proxy model is solved using a multi-objective optimization algorithm. The optimal solutions of different optimization algorithms are obtained, including:

[0134] First, the Utopia point method is used to compare the advantages and disadvantages of different solutions (Pareto solutions obtained by different multi-objective optimization algorithms) to obtain the optimal solution; then, the design variables corresponding to the optimal solution are imported into the thermal simulation analysis model to obtain the maximum temperature T of the electric drive assembly. max , cooling water pressure Pw and cooling oil pressure P o The simulation value of the electric drive assembly is based on the maximum temperature T max , cooling water pressure P w and cooling oil pressure P o The deviation between the optimized value and the simulated value is selected to obtain the maximum temperature T of the electric drive assembly. max , cooling water pressure P w and cooling oil pressure P o The minimum deviation between the optimized value and the simulated value is obtained to obtain the final optimization solution.

[0135] As shown in Table 3, the optimal solutions of the three multi-objective optimization algorithms (NSGA-II, TOP and SparseEA) calculated by the Utopia point method are calculated. In order to verify the feasibility of the optimization results, the design variables corresponding to the optimal solutions (such as oil inner diameter r, nozzle size c, water channel connection width b, oil inlet diameter R, oil inlet depth L, cooling oil flow rate V) are calculated. o and cooling water flow rate V w ) is substituted into the thermal simulation analysis model of the electric drive assembly to obtain the maximum temperature T of the simulated electric drive assembly. max , cooling water pressure P w and cooling oil pressure P o The research objectives are to calculate the maximum temperature T of the electric drive assembly. max , cooling water pressure P w and cooling oil pressure P o The deviation between the optimized value and the simulated value is calculated, and the NSGA-II optimization scheme with the smallest deviation is selected as the final optimization scheme, as shown in Table 4 and Figure 15-19 shown.

[0136] Table 3 Optimal solution sets of different optimization algorithms

[0137]

[0138] Table 4 Comparison of solutions before and after optimization

[0139]

[0140] Refer to Table 4 and Figure 15-19 As shown in the figure, it can be found that by changing the structural parameters of the electric drive assembly and the cooling process variables, the maximum temperature can be reduced by about 5K, especially the temperature of the winding and stator areas has been significantly improved. Although the optimized solution will sacrifice a certain coolant pressure, this is also within the design allowable range. The final results show that the optimization design of the electric drive assembly heat dissipation combined with the electric drive assembly proxy model can significantly improve the heat dissipation effect of the electric drive assembly and provide theoretical support for the production and processing of the actual electric drive assembly heat dissipation system. In addition, since the service life of the induction motor is greatly affected by the maximum operating temperature, improving the internal heat dissipation of the motor can increase the life of the motor.

[0141] ③ Experimental verification of the optimization scheme

[0142] Finally, based on the modeling and optimization results, experimental verification was performed. Considering the processing conditions, some parameters were regularized. At the same time, since the winding temperature is generally the highest temperature of the electric drive assembly, this temperature was used as the verification object. The experimental process is as follows, and the final verification results are as follows Fig. 20 shown.

[0143] a) Before the experiment, the temperature of the electric drive assembly winding is recorded by the host computer;

[0144] b) During the experiment, the electric drive assembly is operated under the rated operating conditions of rated voltage 400V, rated speed 6500rpm, continuous torque 125N·m, and coolant temperature 25℃. The host computer records the winding temperature, continuous torque, and rated speed of the electric drive assembly;

[0145] c) Start when the electric drive assembly outputs a stable rated power of 85kW. Record the temperature every 30 seconds and observe the change in the temperature of the electric drive assembly winding on the host computer until the maximum and minimum values ​​of ten consecutive values ​​are less than 1K;

[0146] d) Draw the temperature rise curve under rated working conditions, such as Fig.21 shown.

[0147] Fig.21 To test the actual temperature change curve of the electric drive assembly, it can be found from the figure that the temperature of the electric drive assembly basically reaches a stable state after about 1000s, indicating that the heat generation and heat dissipation of the electric drive assembly have reached a balance at this time. Because the highest temperature of the electric drive assembly is reflected at the end of the winding, the corresponding temperature sensor is designed at the end of the winding for data acquisition. The experimentally measured maximum steady-state heat dissipation temperature of the electric drive assembly is 393K, which can better meet the heat dissipation requirements of the electric drive assembly. At the same time, the experimental value is only 1.169K different from the 391.831K obtained by design simulation, which shows that the established electric drive assembly thermal analysis model and heat dissipation optimization framework have high reliability.

[0148] In summary, refer to Fig. 22 As shown, Fig. 22 It is a structural schematic diagram of an electric drive assembly heat dissipation optimization framework provided by the present invention; before optimizing and analyzing the heat dissipation method of the electric drive assembly, a corresponding electric drive assembly heat dissipation optimization framework is first constructed, and through the electric drive assembly heat dissipation optimization framework, parametric modeling and optimization of the electric drive assembly heat dissipation system can be effectively realized; wherein, the electric drive assembly heat dissipation optimization framework includes: scheme design and parameter selection, construction of agent model, evaluation and improvement of agent model, optimization model and algorithm solution.

[0149] Specifically, the heat dissipation method of the electric drive assembly is optimized and designed to find the optimal heat dissipation parameters of the electric drive assembly and obtain the best heat dissipation effect. First, the heat dissipation optimization framework of the electric drive assembly is constructed, and then the design parameters are selected according to the heat dissipation optimization framework of the electric drive assembly, and the proxy model is constructed and improved. Then, the best proxy model method is verified and selected. For the constructed proxy model, a multi-objective optimization algorithm is selected for optimization and solution. Finally, the optimization results are analyzed and explained to obtain the optimal design solution.

[0150] In some optional embodiments, the Utopia point method is used to compare the advantages and disadvantages of different optimization algorithms to obtain the actual optimal solution, including:

[0151] The data is normalized, and the target repair weight is given according to the importance of the research objectives (cooling water pressure, cooling oil pressure and maximum temperature of the electric drive assembly); the target repair weight is calculated using the utopian point to obtain the Euclidean distance between the point represented by each optimal solution and the utopian point; the target repair weight is calculated using the dystopian point to obtain the Euclidean distance between the point represented by each optimal solution and the dystopian point; the fitness is obtained based on the Euclidean distance between the point represented by each optimal solution and the utopian point and the Euclidean distance between the point represented by each optimal solution and the dystopian point.

[0152] Specifically, the data (such as cooling water pressure, cooling oil pressure and maximum temperature of the electric drive assembly) are normalized, and target repair weights are given according to the importance of the research objectives (cooling water pressure, cooling oil pressure and maximum temperature of the electric drive assembly).

[0153] Specifically, the calculation method of the normalized value is repair weight * (original value - worst value) / normalized range. The selection of normalized range and repair weight should be determined according to the specific numerical value, numerical range and project requirements. In this design example, the normalized range is the difference between the maximum and minimum values ​​in the Pareto front solution. It can be seen from the distribution range of the Pareto solution that the distribution value of the maximum temperature value is concentrated around 400K. This temperature value is relatively high under rated operating conditions. Therefore, the corresponding repair weight should be increased to precipitate a better temperature optimization result. The numerical values ​​of the cooling oil pressure and the cooling water pressure are close, and the load requirements for the radiator and the pump body are also close. Therefore, the same repair weight can be selected. In summary, the repair weights of the cooling water pressure, cooling oil pressure, and the maximum temperature of the electric drive assembly are selected as 1, 1, and 3. The value of the utopia point is the optimal value of the Pareto solution of each research objective, and it is normalized. The value of the dystopia point is the water pressure difference calculated based on the rated power of the water pump and the calculated flow rate (the maximum water pressure allowed at the rated power); the oil pressure difference calculated based on the rated power of the oil pump and the calculated flow rate (the maximum oil pressure allowed at the rated power); the maximum operating temperature of the motor, and it is normalized.

[0154] The target repair weight is calculated using the Utopia point P+ to obtain the Euclidean distance S+ between the point Ti represented by each optimal solution and the Utopia point. The specific calculation formula is as follows:

[0155] Where S represents the distance, i=1 means the distance starts from unit 1;

[0156] The target repair weight is calculated using the dystopia point P- to obtain the Euclidean distance S- between the point Ti represented by each optimal solution and the dystopia point. The specific calculation formula is as follows:

[0157]

[0158] Where S represents the distance, i=1 means the distance starts from unit 1;

[0159] According to the Euclidean distance between the point represented by each optimal solution and the Utopia point and the Euclidean distance between the point represented by each optimal solution and the Dystopia point, the fitness C is obtained. The fitness C represents the difference between the optimization point and the ideal optimal point. The specific calculation formula is as follows:

[0160]

[0161] The above scheme can be understood as determining the fitness through the Euclidean distance between the point represented by each optimal solution and the utopian point and the Euclidean distance between the point represented by each optimal solution and the dystopian point, and selecting the point with the best fitness as the actual optimal solution.

[0162] It can be seen from the above embodiments that the heat dissipation method of the vehicle electric drive assembly provided by the present invention achieves at least the following beneficial effects:

[0163] The heat dissipation method of the vehicle electric drive assembly provided by the present invention comprises obtaining a sample space with a reasonable distribution of sample points; selecting the maximum temperature, cooling water pressure and cooling oil pressure of the electric drive assembly as research targets according to the sample space with a reasonable distribution of sample points, and constructing an electric drive assembly proxy model; optimizing the electric drive assembly proxy model through an incremental learning strategy to obtain a high-precision proxy model; using the oil circuit structure of the motor as a design variable, and the maximum temperature and coolant pressure of the electric drive assembly as optimization targets, solving the high-precision proxy model using a multi-objective optimization algorithm, and obtaining the optimal solutions of different optimization algorithms. The above scheme is adopted to optimize the electric drive assembly proxy model through an incremental learning strategy, so that the high-precision proxy model is obtained. The accuracy of the model has been significantly improved, and the R-square value in the final evaluation index can reach 0.983, which can truly reflect the actual physical situation; then different multi-objective optimization algorithms are used to solve the high-precision proxy model to obtain the optimal solution of different multi-objective optimization algorithms. The optimal solution of different multi-objective optimization algorithms is obtained through the Utopia point method. It has been verified to have high reliability and can significantly improve the heat dissipation effect of the electric drive assembly, providing theoretical support for the production and processing of the actual electric drive assembly heat dissipation system; in addition, since the service life of the induction motor is greatly affected by the maximum operating temperature of the electric drive assembly, improving the internal heat dissipation of the motor can increase the life of the motor.

[0164] Embodiment 2

[0165] Reference Fig.23 As shown, Fig.23 is a schematic diagram of the structure of a heat dissipation system of a vehicle electric drive assembly provided by the present invention. This embodiment provides a heat dissipation system of a vehicle electric drive assembly, which is applied to the heat dissipation method of the above-mentioned vehicle electric drive assembly, including:

[0166] An acquisition module 100 is used to acquire a sample space with a reasonable distribution of sample points;

[0167] A construction module 200 is used to select the maximum temperature, cooling water pressure and cooling oil pressure of the electric drive assembly as research targets according to a sample space with reasonable sample point distribution, and to construct an electric drive assembly agent model;

[0168] An optimization module 300 is used to optimize the electric drive assembly proxy model through an incremental learning strategy to obtain a high-precision proxy model;

[0169] The calculation module 400 is used to solve the high-precision proxy model using a multi-objective optimization algorithm with the oil circuit structure of the motor as the design variable and the maximum temperature and coolant pressure of the electric drive assembly as the optimization target, so as to obtain the optimal solution of different optimization algorithms.

[0170] Embodiment 3

[0171] This embodiment provides an electronic device, the electronic device comprising:

[0172] memory for storing programs; and

[0173] The processor is used to execute any one of the heat dissipation methods for the vehicle electric drive assembly provided in the present application by calling a program stored in the memory.

[0174] Embodiment 4

[0175] This embodiment provides a computer-readable storage medium, which stores computer instructions, and the computer instructions are used to enable a computer to execute any one of the vehicle electric drive assembly heat dissipation methods provided in this application.

[0176] Although some specific embodiments of the present invention have been described in detail by way of example, it will be appreciated by those skilled in the art that the above examples are for illustration only and are not intended to limit the scope of the present invention. It will be appreciated by those skilled in the art that the above embodiments may be modified without departing from the scope and spirit of the present invention. The scope of the present invention is defined by the appended claims.

Claims

1. A heat dissipation method for a vehicle electric drive assembly, It is characterized in that include: Obtain a sample space with reasonable distribution of sample points; According to the sample space with reasonable distribution of the sample points, the maximum temperature, cooling water pressure and cooling oil pressure of the electric drive assembly are selected as research targets, and the proxy model of the electric drive assembly is constructed; The electric drive assembly agent model is optimized by an incremental learning strategy to obtain a high-precision agent model; Taking the oil circuit structure of the motor as the design variable, and the maximum temperature and coolant pressure of the electric drive assembly as the optimization targets, the high-precision proxy model is solved using a multi-objective optimization algorithm to obtain the optimal solutions of different optimization algorithms.

2. The heat dissipation method of the vehicle electric drive assembly according to claim 1, It is characterized in that The oil circuit structure of the motor is used as the design variable, the maximum temperature and coolant pressure of the electric drive assembly are used as the optimization targets, and the high-precision proxy model is solved by using a multi-objective optimization algorithm. After obtaining the optimal solutions of different optimization algorithms, the following steps are included: Use the Utopia point method to compare the advantages and disadvantages of different optimization algorithms and obtain the actual optimal solution; Importing the design variables corresponding to the optimal solution into the thermal simulation analysis model to obtain simulation values ​​of peak torque, cooling water pressure and cooling oil pressure; According to the deviation between the optimized value and the simulated value of the maximum temperature of the electric drive assembly, the cooling water pressure and the cooling oil pressure, the minimum deviation between the optimized value and the simulated value of the peak torque, the cooling water pressure and the cooling oil pressure is selected to obtain the final optimization solution.

3. The heat dissipation method of the vehicle electric drive assembly according to claim 2, It is characterized in that The above-mentioned Utopia point method is used to compare the advantages and disadvantages of different optimization algorithms, and the actual optimal solution is obtained, including: Normalize the data and give target restoration weights according to the importance of the research target; The target restoration weight is calculated using the Utopia point to obtain the Euclidean distance between the point represented by each optimal solution and the Utopia point; The target restoration weight is calculated using the dystopia point to obtain the Euclidean distance between the point represented by each optimal solution and the dystopia point; The fitness is obtained according to the Euclidean distance between the point represented by each optimal solution and the Utopia point and the Euclidean distance between the point represented by each optimal solution and the Dystopia point.

4. The heat dissipation method of a vehicle electric drive assembly according to claim 1, It is characterized in that The electric drive assembly agent model is optimized by the incremental learning strategy to obtain a high-precision agent model including: Using evaluation indicators to evaluate the accuracy of the electric drive assembly proxy model; The incremental Kriging method is used to optimize the design variables with insufficient precision in the evaluated electric drive assembly proxy model, and the high-precision proxy model is obtained by continuously optimizing and updating the initial training set through the expected improvement method.

5. The heat dissipation method of the vehicle electric drive assembly according to claim 4, It is characterized in that The method of optimizing the design variables with insufficient precision in the evaluated electric drive assembly proxy model by using the incremental kriging method, and continuously optimizing and updating the initial training set by the expectation improvement method to obtain the high-precision proxy model includes: Assuming that the number of new sample points added is s, the incremental Kriging expression is: Among them, the relevant update process variables are as follows: The expected improvement method is used to increase sample points in the incremental kriging method. Assume that for any design variable x, the response value y is a random variable that obeys a normal distribution, and its mean is Its probability density function is: Assume that the optimal response value of the current sample is y b , then the expected value of the objective function at this design point is: The expected improved function is as follows: where φ is the cumulative distribution function of the standard normal distribution.

6. The heat dissipation method of a vehicle electric drive assembly according to claim 1, It is characterized in that The step of obtaining the sample space where the sample points are reasonably distributed comprises: Obtaining heat dissipation design variables corresponding to the electric drive assembly, wherein the heat dissipation design variables include oil inner diameter, nozzle size, water channel connection width, oil inlet diameter, oil inlet depth, cooling oil flow rate, and cooling water flow rate; The heat dissipation design variables are sampled using an experimental design method, and a sample space in which the sample points are reasonably distributed is selected.

7. A cooling system for a vehicle electric drive assembly, It is characterized in that A heat dissipation method for a vehicle electric drive assembly according to any one of claims 1 to 6, comprising: An acquisition module is used to obtain a sample space with a reasonable distribution of sample points; A construction module is used to select the maximum temperature, cooling water pressure and cooling oil pressure of the electric drive assembly as research targets according to the sample space with reasonable distribution of the sample points, and to construct an agent model of the electric drive assembly; An optimization module is used to optimize the electric drive assembly proxy model through an incremental learning strategy to obtain a high-precision proxy model; The calculation module is used to solve the high-precision proxy model using a multi-objective optimization algorithm with the oil circuit structure of the motor as the design variable and the maximum temperature and coolant pressure of the electric drive assembly as the optimization target, so as to obtain the optimal solution of different optimization algorithms.

8. An electronic device, It is characterized in that The electronic device comprises: memory for storing programs; and A processor, configured to execute the heat dissipation method for a vehicle electric drive assembly as described in any one of claims 1 to 6 by calling the program stored in the memory.

9. A computer-readable storage medium, It is characterized in that The computer-readable storage medium stores computer instructions, and the computer instructions are used to enable the computer to execute the heat dissipation method for a vehicle electric drive assembly according to any one of claims 1 to 6.

Citation Information

Patent Citations

  • Incremental learning-based design optimization method for oil-water mixed heat dissipation system of induction motor

    CN114861556A

  • Permanent magnet synchronous motor cooling system optimization method based on error compensation

    CN116361906A

  • Heat sink system

    US20220117109A1