Parameter optimization method and device, computer equipment, storage medium and program product
By using radial-based neural networks and genetic algorithms to optimize the parameters of the vehicle's energy consumption model in vehicle energy consumption analysis, the problem of inaccurate prediction of energy consumption data in traditional technology is solved, and the accuracy of energy consumption analysis is improved.
Patent Information
- Application Number
- CN202411154356.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-21
- Publication Date
- 2024-12-13
AI Technical Summary
The model structure and parameters of the traditional vehicle energy consumption analysis model have not been optimized, resulting in low accuracy of energy consumption data prediction.
The parameters of the vehicle's energy consumption mechanism model are optimized by radial-based neural network combined with genetic algorithms. The sample vehicle data under different operating conditions are input and processed, and the model parameters of the radial-based neural network are optimized to improve the energy consumption data. accuracy.
By optimizing the parameters of the vehicle's energy consumption mechanism model, the prediction accuracy of vehicle energy consumption data is improved and the model's energy consumption analysis ability under different working conditions is enhanced.
Smart Images
Figure CN119129391A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of artificial intelligence technology, and in particular to a parameter optimization method, apparatus, computer equipment, storage medium and program product. Background Art
[0002] Vehicle energy consumption analysis aims to understand the energy consumption of a vehicle under specific working conditions, including fuel consumption (for traditional fuel vehicles) or electricity consumption (for electric vehicles). Through analysis, high energy consumption links can be identified, providing data support for energy conservation and emission reduction, optimizing vehicle design, and reducing operating costs.
[0003] In traditional technology, an energy consumption analysis model is usually used to obtain vehicle energy consumption data. However, the model structure and model parameters of the energy consumption analysis model are not optimized, and the model accuracy is not high, which leads to low accuracy of the predicted energy consumption data. Summary of the invention
[0004] Based on this, it is necessary to provide a parameter optimization method, device, computer equipment, storage medium and program product to improve the accuracy of predicted energy consumption data in response to the above technical problems.
[0005] In a first aspect, the present application provides a parameter optimization method, the method comprising:
[0006] The sample vehicle data under different operating conditions are input into the radial basis function neural network to obtain the predicted energy consumption data under different operating conditions;
[0007] According to the reference energy consumption data and the predicted energy consumption data under different operating conditions, the model parameters of the radial basis neural network are optimized by using a genetic algorithm; wherein the model parameters are associated with the parameters to be corrected of the whole vehicle energy consumption mechanism model of the vehicle, and the whole vehicle energy consumption mechanism model is used to predict the whole vehicle energy consumption data of the vehicle;
[0008] Based on the optimized model parameters, the parameters to be corrected are optimized.
[0009] In one embodiment, the vehicle energy consumption mechanism model is established in the following manner:
[0010] Obtain the vehicle's motor parameters, battery parameters, and electrical accessory parameters;
[0011] Constructing a drive motor mechanism model according to the motor parameters, constructing a power battery mechanism model according to the battery parameters, and constructing an electric accessory system fitting model according to the electric accessory parameters;
[0012] Based on the mechanism fusion model, a coupling relationship and constraint conditions are established among the drive motor mechanism model, the power battery mechanism model and the electric accessory system fitting model;
[0013] The whole vehicle energy consumption mechanism model is established based on the drive motor mechanism model, the power battery mechanism model, the electric accessory system fitting model, the coupling relationship and the constraint conditions.
[0014] In one embodiment, the sample vehicle data under different operating conditions are obtained by:
[0015] Acquiring candidate vehicle data of the vehicle under different operating conditions;
[0016] The candidate vehicle data is sampled using an optimal Latin hypercube algorithm to obtain sample vehicle data of the vehicle under different operating conditions.
[0017] In one of the embodiments, the radial basis neural network includes an input layer, a radial basis layer and an output layer;
[0018] The sample vehicle data under different operating conditions are input into the radial basis function neural network to obtain the predicted energy consumption data under different operating conditions, including:
[0019] For each operating condition, inputting sample vehicle data under the operating condition into the input layer to extract basic features to obtain basic feature data;
[0020] Inputting the basic characteristic data into the radial base layer for nonlinear transformation processing to obtain nonlinear characteristic data;
[0021] The nonlinear characteristic data is input into the output layer for weighted summation processing to obtain the predicted energy consumption data under the operating condition.
[0022] In one embodiment, the model parameters include the number of nodes and radial basis function centers of the radial base layer, and the weight of the output layer;
[0023] The method of optimizing the model parameters of the radial basis function neural network by using a genetic algorithm according to the reference energy consumption data and the predicted energy consumption data under different operating conditions includes:
[0024] According to the difference between the reference energy consumption data and the predicted energy consumption data under different operating conditions, a genetic algorithm is used to optimize the number of nodes and the center of the radial basis function of the radial base layer and the weight of the output layer.
[0025] In one embodiment, the radial basis function is any one of a Gaussian function, a polynomial function or an inverse multiquadratic function.
[0026] In a second aspect, the present application further provides a parameter optimization device, the device comprising:
[0027] An acquisition module is used to input sample vehicle data under different operating conditions into a radial basis function neural network to obtain predicted energy consumption data under different operating conditions;
[0028] A first optimization module is used to optimize the model parameters of the radial basis neural network using a genetic algorithm according to the reference energy consumption data and the predicted energy consumption data under different operating conditions; wherein the model parameters are associated with the parameters to be corrected of the whole vehicle energy consumption mechanism model of the vehicle, and the whole vehicle energy consumption mechanism model is used to predict the whole vehicle energy consumption data of the vehicle;
[0029] The second optimization module is used to optimize the parameters to be corrected based on the optimized model parameters.
[0030] In a third aspect, the present application further provides a computer device, the computer device comprising a memory and a processor, the memory storing a computer program, and the processor implementing the following steps when executing the computer program:
[0031] The sample vehicle data under different operating conditions are input into the radial basis function neural network to obtain the predicted energy consumption data under different operating conditions;
[0032] According to the reference energy consumption data and the predicted energy consumption data under different operating conditions, the model parameters of the radial basis neural network are optimized by using a genetic algorithm; wherein the model parameters are associated with the parameters to be corrected of the whole vehicle energy consumption mechanism model of the vehicle, and the whole vehicle energy consumption mechanism model is used to predict the whole vehicle energy consumption data of the vehicle;
[0033] Based on the optimized model parameters, the parameters to be corrected are optimized.
[0034] In a fourth aspect, the present application further provides a computer-readable storage medium, wherein the computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the following steps are implemented:
[0035] The sample vehicle data under different operating conditions are input into the radial basis function neural network to obtain the predicted energy consumption data under different operating conditions;
[0036] According to the reference energy consumption data and the predicted energy consumption data under different operating conditions, the model parameters of the radial basis neural network are optimized by using a genetic algorithm; wherein the model parameters are associated with the parameters to be corrected of the whole vehicle energy consumption mechanism model of the vehicle, and the whole vehicle energy consumption mechanism model is used to predict the whole vehicle energy consumption data of the vehicle;
[0037] Based on the optimized model parameters, the parameters to be corrected are optimized.
[0038] In a fifth aspect, the present application further provides a computer program product, the computer program product comprising a computer program, which implements the following steps when executed by a processor:
[0039] The sample vehicle data under different operating conditions are input into the radial basis function neural network to obtain the predicted energy consumption data under different operating conditions;
[0040] According to the reference energy consumption data and the predicted energy consumption data under different operating conditions, the model parameters of the radial basis neural network are optimized by using a genetic algorithm; wherein the model parameters are associated with the parameters to be corrected of the whole vehicle energy consumption mechanism model of the vehicle, and the whole vehicle energy consumption mechanism model is used to predict the whole vehicle energy consumption data of the vehicle;
[0041] Based on the optimized model parameters, the parameters to be corrected are optimized.
[0042] The above parameter optimization method, device, computer equipment, storage medium and program product input sample vehicle data under different operating conditions into the radial basis neural network to obtain predicted energy consumption data under different operating conditions; and according to the reference energy consumption data and predicted energy consumption data under different operating conditions, the model parameters of the radial basis neural network are optimized by using a genetic algorithm; the model parameters are associated with the parameters to be corrected of the vehicle's whole vehicle energy consumption mechanism model, and the whole vehicle energy consumption mechanism model is used to predict the vehicle's whole vehicle energy consumption data; thus, by optimizing the model parameters of the radial basis neural network, the parameters to be corrected of the whole vehicle energy consumption mechanism model are optimized, and specifically, the parameters to be corrected can be optimized based on the optimized model parameters. The above scheme optimizes the model parameters of the radial basis neural network according to the association between the parameters to be corrected of the vehicle's whole vehicle energy consumption mechanism model and the model parameters, and then determines the parameters of the new whole vehicle energy consumption mechanism model according to the optimized model parameters and the association, thereby optimizing the parameters to be corrected of the whole vehicle energy consumption mechanism model, thereby improving the accuracy of the predicted energy consumption data. BRIEF DESCRIPTION OF THE DRAWINGS
[0043] Figure 1 A diagram showing an application environment of a parameter optimization method in an embodiment;
[0044] Figure 2 A schematic diagram of optimizing a vehicle energy consumption mechanism model in an embodiment;
[0045] Figure 3 A schematic diagram of a flow chart of a parameter optimization method in one embodiment;
[0046] Figure 4 A schematic diagram of a process for establishing a vehicle energy consumption mechanism model in an embodiment;
[0047] Figure 5 A schematic diagram of a process for obtaining sample vehicle data in one embodiment;
[0048] Figure 6 A schematic diagram of a process for obtaining predicted energy consumption data in one embodiment;
[0049] Figure 7 A schematic flow chart of another parameter optimization method in one embodiment;
[0050] Figure 8 is a structural block diagram of a parameter optimization device in one embodiment;
[0051] Fig. 9 FIG. 4 is a diagram showing the internal structure of a computer device in one embodiment. DETAILED DESCRIPTION
[0052] In order to make the purpose, technical solution and advantages of the present application more clearly understood, the present application is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application.
[0053] The parameter optimization method provided in the embodiment of the present application can be applied to Figure 1 In the application environment shown, it is specifically used in the application scenario of optimizing the parameters to be corrected of the vehicle energy consumption mechanism model. Among them, the vehicle energy consumption mechanism model can simulate and analyze the energy consumption of the car under different working conditions, including driving energy consumption, battery loss, auxiliary system energy consumption and charging loss. The vehicle energy consumption mechanism model comprehensively reflects the energy consumption characteristics of the car by considering the various components of the vehicle power system (such as batteries, motors, transmission systems, auxiliary systems, etc.) and their interactions.
[0054] In order to improve the accuracy of the vehicle energy consumption data simulated by the vehicle energy consumption mechanism model, it is necessary to optimize the vehicle energy consumption mechanism model. Figure 1 As shown, the multi-dimensional operating condition data of the vehicle can be collected, and the energy consumption mechanism model of the whole vehicle can be established based on the collected multi-dimensional operating condition data; then the parameters to be corrected of the energy consumption mechanism model of the whole vehicle can be transmitted to the neural network model, and the preset machine learning algorithm can be used to optimize the parameters to be corrected of the energy consumption mechanism model of the whole vehicle based on the collected multi-dimensional operating condition data as the training set and the verification set.
[0055] For details, see Figure 2 , Figure 2A schematic diagram for optimizing a whole vehicle energy consumption mechanism model is provided. Among them, the correlation between the parameters to be corrected of the whole vehicle energy consumption mechanism model and the model parameters of the radial basis neural network can be established in advance. Then, the parameters to be corrected of the whole vehicle energy consumption mechanism model can be transmitted to the radial basis neural network, and the radial basis neural network is used in the embodiment of the present application to optimize the whole vehicle energy consumption mechanism model. The optimal Latin hypercube algorithm can be used to sample the collected multi-dimensional operating condition data to obtain sample point data. Then, according to the sample point data, a genetic algorithm can be used to determine the error according to the difference between the predicted energy consumption data and the reference energy consumption data in the sample point data; and based on the error, the model parameters of the radial basis neural network are optimized, thereby optimizing the parameters to be corrected of the whole vehicle energy consumption mechanism model; and the optimized parameters to be corrected are used to update the whole vehicle energy consumption mechanism model, thereby optimizing the whole vehicle energy consumption mechanism model.
[0056] The parameter optimization method provided in the embodiment of the present application can be executed by a server or by a terminal with a certain computing power. Among them, the server can be implemented by an independent server or a server cluster composed of multiple servers. The terminal can be, but is not limited to, various personal computers, laptops, smart phones, tablet computers, Internet of Things devices and portable wearable devices. The Internet of Things devices can be smart speakers, smart TVs, smart air conditioners, smart car-mounted devices, etc. Portable wearable devices can be smart watches, smart bracelets, head-mounted devices, etc.
[0057] In one embodiment, Figure 3 As shown, a parameter optimization method is provided, which is described by taking the method applied to a server as an example, and includes the following steps:
[0058] S301, inputting sample vehicle data under different operating conditions into a radial basis function neural network to obtain predicted energy consumption data under different operating conditions.
[0059] Exemplarily, sample vehicle data under different operating conditions may be acquired in advance, wherein the sample vehicle data includes but is not limited to basic vehicle parameters, power system parameters, driving condition data, and auxiliary system parameters; wherein the basic vehicle parameters may include gross mass, vehicle model, wheel diameter, and body size, etc. Power system parameters may include battery parameters, motor parameters, and transmission system parameters, etc. Driving condition data may include speed curves, acceleration / deceleration data, and driving environment, etc. Auxiliary system parameters may include air conditioning system parameters, electric power steering system parameters, and vehicle electrical parameters.
[0060] The correlation between the parameters to be corrected of the whole vehicle energy consumption model and the model parameters of the radial basis neural network can be established in advance, and then the parameters to be corrected of the whole vehicle energy consumption model can be optimized indirectly by optimizing the model parameters of the radial basis neural network.
[0061] Specifically, sample vehicle data under different operating conditions can be input into a radial basis function neural network, and the radial basis function neural network is used to predict the energy consumption data of the vehicle to obtain predicted energy consumption data under different operating conditions.
[0062] S302, optimizing the model parameters of the radial basis function neural network using a genetic algorithm according to the reference energy consumption data and the predicted energy consumption data under different operating conditions.
[0063] Furthermore, the model parameters are associated with the parameters to be corrected of the vehicle's whole vehicle energy consumption mechanism model, and the whole vehicle energy consumption mechanism model is used to predict the vehicle's whole vehicle energy consumption data. The reference energy consumption data is the actual vehicle energy consumption data corresponding to the collected sample vehicle data, so the model parameters of the radial basis neural network can be optimized using a genetic algorithm based on the reference energy consumption data and the predicted energy consumption data under different operating conditions. The genetic algorithm mainly optimizes the model parameters of the radial basis neural network by simulating natural selection and genetic mechanisms to search for the optimal parameter combination, so as to improve the performance of the radial basis neural network.
[0064] S303, optimizing the parameters to be corrected based on the optimized model parameters.
[0065] Exemplarily, based on the correlation between the model parameters and the parameters to be corrected of the vehicle energy consumption mechanism model, the optimized parameters to be corrected can be obtained according to the optimized model parameters, thereby achieving optimization of the parameters to be corrected.
[0066] Furthermore, a new vehicle energy consumption mechanism model can be constructed based on the optimized parameters to be corrected, so that the constructed vehicle energy consumption model has higher prediction accuracy, thereby improving the accuracy of the acquired vehicle energy consumption data.
[0067] The above parameter optimization method inputs sample vehicle data under different operating conditions into the radial basis neural network to obtain predicted energy consumption data under different operating conditions; and according to the reference energy consumption data and predicted energy consumption data under different operating conditions, the model parameters of the radial basis neural network are optimized by using a genetic algorithm; the model parameters are associated with the parameters to be corrected of the vehicle's whole vehicle energy consumption mechanism model, and the whole vehicle energy consumption mechanism model is used to predict the vehicle's whole vehicle energy consumption data; thus, by optimizing the model parameters of the radial basis neural network, the parameters to be corrected of the whole vehicle energy consumption mechanism model are optimized, and specifically, the parameters to be corrected can be optimized based on the optimized model parameters. The above scheme optimizes the model parameters of the radial basis neural network according to the association between the parameters to be corrected of the vehicle's whole vehicle energy consumption mechanism model and the model parameters, and then determines the parameters of the new whole vehicle energy consumption mechanism model according to the optimized model parameters and the association, thereby optimizing the parameters to be corrected of the whole vehicle energy consumption mechanism model, thereby improving the accuracy of the predicted energy consumption data.
[0068] In some optional implementations, see Figure 4 , Figure 4 A schematic diagram of a process for establishing a vehicle energy consumption mechanism model is provided, which specifically includes the following steps:
[0069] S401, obtaining motor parameters, battery parameters and electrical accessory parameters of the vehicle.
[0070] Exemplarily, the vehicle's motor parameters include but are not limited to the motor's power, torque, efficiency, etc.; battery parameters include but are not limited to battery capacity, voltage, internal resistance, etc.; electrical accessories include an electrical accessory system including on-board air conditioning and low-voltage electrical accessories, and electrical accessory parameters include but are not limited to the air conditioning's cooling / heating power, energy consumption, and power consumption of on-board electrical appliances such as audio and lighting.
[0071] For example, the motor parameters of the permanent magnet synchronous motor can be extracted based on the motor design theory and loss mathematical model, and the mutual coupling of the electric field, magnetic field, thermal field and mechanical field can be analyzed from the perspective of the power loss of the permanent magnet synchronous motor; the battery parameters of the lithium-ion power battery can also be extracted based on the electrochemical theoretical mathematical model; the electrical accessory parameters can also be extracted according to the energy consumption composition of the electrical accessory system, and the energy consumption expression of the air-conditioning system and the low-voltage electrical accessory system can be constructed based on the electrical accessory parameters.
[0072] S402, constructing a drive motor mechanism model according to motor parameters, constructing a power battery mechanism model according to battery parameters, and constructing an electric accessory system fitting model according to electric accessory parameters.
[0073] For example, according to the motor parameters, MATLAB / Simulink software can be used to establish a multi-field coupled dynamic simulation model of a permanent magnet synchronous motor, simulate the performance output, dynamic characteristics and working efficiency of the permanent magnet synchronous motor, and realize the mapping of the design parameters of the permanent magnet synchronous motor to the performance indicators. COMSOL Multiphysics multi-physics field simulation software can be used to build an electrochemical model of the lithium battery cycle condition, and the battery open circuit voltage, positive and negative electrode potential, battery internal resistance, precise battery SOC and other performance of the battery during the cycle condition can be simulated and analyzed. Based on the sample vehicle energy consumption data set, the least squares method can also be used to analyze the inherent complex nonlinear relationship of the sample vehicle energy consumption data set, fit and extract the unknown parameters in the energy consumption expression of the air-conditioning system and the low-voltage electric accessory system, and identify the energy consumption system parameters of the electric accessories, so as to realize the mapping of parameters such as ambient temperature and air-conditioning gear to the energy consumption of the electric accessories.
[0074] S403, based on the mechanism fusion model, establishing the coupling relationship and constraint conditions among the drive motor mechanism model, the power battery mechanism model and the electric accessory system fitting model.
[0075] Exemplarily, the mechanism fusion model is used to fuse the drive motor mechanism model, the power battery mechanism model and the electric accessory system fitting model, and construct the coupling relationship and constraint conditions between the drive motor mechanism model, the power battery mechanism model and the electric accessory system fitting model according to the motor parameters simulated by the drive motor mechanism model, the battery parameters simulated by the power battery mechanism model, and the electric accessory parameters fitted by the electric accessory system fitting model.
[0076] S404: Establish a vehicle energy consumption mechanism model based on the drive motor mechanism model, the power battery mechanism model, the electric accessory system fitting model, the coupling relationship and the constraint conditions.
[0077] Exemplarily, based on the drive motor mechanism model, power battery mechanism model, electric accessory system fitting model, coupling relationship and constraint conditions, the drive motor mechanism model, power battery mechanism model and electric accessory system fitting model can be integrated to obtain the vehicle energy consumption mechanism model.
[0078] In an embodiment of the present application, an implementable method for constructing a whole vehicle energy consumption mechanism model is provided, which can realize the mapping between motor parameters, battery parameters and electrical accessory parameters to whole vehicle energy consumption data, and then the whole vehicle energy consumption data of the vehicle can be efficiently obtained according to the whole vehicle energy consumption mechanism model.
[0079] In some optional implementations, in the process of optimizing the model parameters of the radial basis function neural network in the above embodiments, it is necessary to feed sample vehicle data into the radial basis function neural network. In order to obtain sample vehicle data under more multidimensional working conditions, sampling can be performed from a large number of candidate vehicle data to obtain sample vehicle data under multidimensional working conditions.
[0080] See also Figure 5 , Figure 5 A schematic diagram of a process for obtaining sample vehicle data is provided, which specifically includes the following steps:
[0081] S501, obtaining candidate vehicle data of the vehicle under different operating conditions.
[0082] Exemplarily, the candidate vehicle data may be vehicle data collected when the vehicle is in different operating conditions, and the collected vehicle data includes basic vehicle parameters, power system parameters, driving condition data, auxiliary system parameters and vehicle energy consumption data.
[0083] S502, using the optimal Latin hypercube algorithm to sample candidate vehicle data to obtain sample vehicle data under different operating conditions.
[0084] Exemplarily, in order to obtain sample vehicle data under more dimensional working conditions, the optimal Latin hypercube algorithm can be used to sample candidate vehicle data. The optimal Latin hypercube algorithm determines the value range of each input variable by analyzing the probability distribution of the input variables. Then, according to these probability distributions, the initial sampling points are randomly generated within the value range. Through the iterative optimization algorithm, the positions of the sampling points are continuously adjusted to maximize the spatial filling of the sampling points and minimize the correlation between the input variables. Finally, a set of optimal sampling points that are evenly distributed in space and have low correlation with each other are obtained. The optimal Latin hypercube algorithm is a hierarchical Monte Carlo sampling method suitable for uniform sampling in multi-dimensional space. In the embodiment of the present application, it is used to collect sample points of the variables of the energy consumption mechanism model of the whole vehicle. In this way, sample vehicle data of the vehicle under different operating conditions can be obtained.
[0085] In an embodiment of the present application, a method for obtaining sample vehicle data is provided, which can enable the obtained sample vehicle data to cover more dimensional working conditions, thereby enabling the radial basis function neural network to learn more dimensional feature data, thereby improving the model prediction accuracy.
[0086] In some optional implementations, the radial basis neural network may include an input layer, a radial basis layer, and an output layer. The transformation from the input layer to the radial basis layer is nonlinear, while the transformation from the radial basis layer to the output layer is linear. The basic idea of the radial basis neural network is to use the radial basis as the "basis" of the neural unit to form the radial basis layer, so that the input vector can be directly mapped to the radial basis layer without connecting through weights. The center point of the radial basis function of each neuron corresponds to the input of each sampled data. When the center point of the radial basis is determined, this mapping relationship is also determined. The mapping from the radial basis layer to the output layer is linear, that is, the output of the network is the linear weighted sum of the output of the neural unit, and the weight here is the network adjustable parameter. Among them, the role of the radial basis layer is to map the vector from a low dimension to a high dimension, so that the low-dimensional linearly inseparable situation can become linearly separable in the high dimension, which is mainly the idea of the kernel function. In this way, the mapping from input to output of the network is nonlinear, while the network output is linear with respect to the adjustable parameters. The weight of the network can be directly solved by a system of linear equations, thereby greatly accelerating the learning speed and avoiding the local minimum problem.
[0087] Based on this, see Figure 6 , Figure 6 A schematic diagram of a process for obtaining predicted energy consumption data is provided, which specifically includes the following steps:
[0088] S601, for each operating condition, inputting sample vehicle data under the operating condition into the input layer to extract basic features to obtain basic feature data.
[0089] Exemplarily, for each operating condition, sample vehicle data under the operating condition may be input into the input layer to extract basic features to obtain basic feature data, and then the input layer transmits the basic feature data to the radial base layer.
[0090] S602, inputting the basic characteristic data into the radial base layer for nonlinear transformation processing to obtain nonlinear characteristic data.
[0091] Exemplarily, the neurons of the radial base layer use radial basis function as activation function. The radial basis function is a distance-based function that can describe the similarity between feature data. The basic feature data can be input into the radial base layer, and the basic feature data can be nonlinearly transformed using the radial basis function to obtain nonlinear feature data. The radial basis function can be any one of a Gaussian function, a polynomial function, or an inverse multi-quadratic function.
[0092] S603, inputting the nonlinear characteristic data into the output layer for weighted summation processing to obtain predicted energy consumption data under operating conditions.
[0093] For example, the output layer is usually a simple linear function, which is responsible for linearly combining the output of the radial base layer to obtain the predicted energy consumption data; the weight of the output layer needs to be trained through a learning algorithm to minimize the error between the output predicted energy consumption data and the reference energy consumption data. In this way, the nonlinear feature data can be input into the output layer for weighted summation processing to obtain the predicted energy consumption data under the operating conditions.
[0094] The embodiments of the present application provide a structure of a radial basis function neural network and an achievable method for obtaining predicted energy consumption data based on the radial basis function neural network, providing a basis for optimizing the model parameters of the radial basis function neural network.
[0095] Further, based on the structure of the radial basis neural network provided in the above embodiment, the model parameters of the radial basis neural network include the number of nodes and the center of the radial basis function of the radial base layer, and the weight of the output layer, and the specific process of optimizing the model parameters of the radial basis neural network is as follows:
[0096] The correlation between the parameters to be corrected of the whole vehicle energy consumption mechanism model and the number of nodes and the center of the radial basis function of the radial base layer, as well as the weight of the output layer can be established in advance. Then, according to the difference between the reference energy consumption data and the predicted energy consumption data under different operating conditions, the number of nodes and the center of the radial basis function of the radial base layer, as well as the weight of the output layer can be optimized by a genetic algorithm, thereby optimizing the model parameters of the radial basis neural network. Then, according to the correlation, the parameters to be corrected of the new whole vehicle energy consumption mechanism model can be obtained.
[0097] Among them, genetic algorithms are a type of search algorithm used to solve optimization problems. Genetic algorithms are a family of search algorithms inspired by natural evolution theory. By imitating the process of natural selection and reproduction, genetic algorithms can provide high-quality solutions to a variety of problems involving search, optimization, and learning. At the same time, they are similar to natural evolution and can therefore overcome some of the obstacles encountered by traditional search and optimization algorithms, especially for problems with a large number of parameters and complex mathematical representations.
[0098] In an embodiment of the present application, a specific implementation process for optimizing the model parameters of a radial basis neural network is provided, by optimizing the number of nodes of the radial base layer and the center of the radial basis function, as well as the weight of the output layer, thereby optimizing the parameters to be corrected of the energy consumption mechanism model of the whole vehicle.
[0099] In some optional implementations, see Figure 7 , Figure 7 A flow chart of another parameter optimization method is provided, which specifically includes the following steps:
[0100] S701, obtaining candidate vehicle data of the vehicle under different operating conditions.
[0101] S702, sampling the candidate vehicle data using the optimal Latin hypercube algorithm to obtain sample vehicle data of the vehicle under different operating conditions.
[0102] S703, inputting sample vehicle data under different operating conditions into the input layer to extract basic features, and obtaining basic feature data under different operating conditions.
[0103] S704, inputting the basic characteristic data under different operating conditions into the radial base layer for nonlinear transformation processing to obtain the nonlinear characteristic data under different operating conditions.
[0104] S705, inputting the nonlinear characteristic data under different operating conditions into the output layer for weighted summation processing to obtain predicted energy consumption data under different operating conditions.
[0105] S706, according to the difference between the reference energy consumption data and the predicted energy consumption data under different operating conditions, a genetic algorithm is used to optimize the number of nodes of the radial base layer and the center of the radial basis function, as well as the weight of the output layer.
[0106] S707, optimizing the parameters to be corrected based on the optimized model parameters.
[0107] S708: construct a new vehicle energy consumption mechanism model based on the optimized parameters to be corrected.
[0108] The specific process of the above S701 to S708 can refer to the description of the above method embodiment, and its implementation principle and technical effect are similar, which will not be repeated here.
[0109] It should be understood that, although the various steps in the flowcharts involved in the above-mentioned embodiments are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence according to the order indicated by the arrows. Unless there is a clear explanation in this article, the execution of these steps does not have a strict order restriction, and these steps can be executed in other orders. Moreover, at least a part of the steps in the flowcharts involved in the above-mentioned embodiments can include multiple steps or multiple stages, and these steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a part of the steps or stages in other steps.
[0110] Based on the same inventive concept, the embodiment of the present application also provides a parameter optimization device for implementing the parameter optimization method involved above. The implementation scheme for solving the problem provided by the device is similar to the implementation scheme recorded in the above method, so the specific limitations in one or more parameter optimization device embodiments provided below can refer to the limitations on the parameter optimization method above, and will not be repeated here.
[0111] In one embodiment, Figure 8 As shown, a parameter optimization device is provided, comprising:
[0112] An acquisition module 10 is used to input sample vehicle data under different operating conditions into a radial basis function neural network to obtain predicted energy consumption data under different operating conditions;
[0113] The first optimization module 20 is used to optimize the model parameters of the radial basis neural network using a genetic algorithm according to the reference energy consumption data and the predicted energy consumption data under different operating conditions; wherein the model parameters are associated with the parameters to be corrected of the whole vehicle energy consumption mechanism model of the vehicle, and the whole vehicle energy consumption mechanism model is used to predict the whole vehicle energy consumption data of the vehicle;
[0114] The second optimization module 30 is used to optimize the parameters to be corrected based on the optimized model parameters.
[0115] The above parameter optimization device inputs sample vehicle data under different operating conditions into the radial basis neural network to obtain predicted energy consumption data under different operating conditions; and according to the reference energy consumption data and predicted energy consumption data under different operating conditions, the model parameters of the radial basis neural network are optimized by using a genetic algorithm; the model parameters are associated with the parameters to be corrected of the vehicle's whole vehicle energy consumption mechanism model, and the whole vehicle energy consumption mechanism model is used to predict the vehicle's whole vehicle energy consumption data; thus, by optimizing the model parameters of the radial basis neural network, the parameters to be corrected of the whole vehicle energy consumption mechanism model are optimized, and specifically, the parameters to be corrected can be optimized based on the optimized model parameters. The above scheme optimizes the model parameters of the radial basis neural network according to the association between the parameters to be corrected of the vehicle's whole vehicle energy consumption mechanism model and the model parameters, and then determines the parameters of the new whole vehicle energy consumption mechanism model according to the optimized model parameters and the association, thereby optimizing the parameters to be corrected of the whole vehicle energy consumption mechanism model, thereby improving the accuracy of the predicted energy consumption data.
[0116] In one embodiment, the vehicle energy consumption mechanism model is established in the following manner:
[0117] Acquire the motor parameters, battery parameters and electric accessory parameters of the vehicle; construct a drive motor mechanism model according to the motor parameters, construct a power battery mechanism model according to the battery parameters, and construct an electric accessory system fitting model according to the electric accessory parameters; based on the mechanism fusion model, establish the coupling relationship and constraint conditions among the drive motor mechanism model, the power battery mechanism model and the electric accessory system fitting model; based on the drive motor mechanism model, the power battery mechanism model, the electric accessory system fitting model, the coupling relationship and the constraint conditions, establish the whole vehicle energy consumption mechanism model.
[0118] In one embodiment, the sample vehicle data under different operating conditions are obtained by:
[0119] Acquire candidate vehicle data of the vehicle under different operating conditions; and sample the candidate vehicle data using an optimal Latin hypercube algorithm to obtain sample vehicle data of the vehicle under different operating conditions.
[0120] In one embodiment, the radial basis neural network includes an input layer, a radial basis layer and an output layer; the acquisition module 10 is specifically used for:
[0121] For each operating condition, the sample vehicle data under the operating condition is input into the input layer for basic feature extraction to obtain basic feature data; the basic feature data is input into the radial base layer for nonlinear transformation processing to obtain nonlinear feature data; the nonlinear feature data is input into the output layer for weighted summation processing to obtain predicted energy consumption data under the operating condition.
[0122] In one embodiment, the model parameters include the number of nodes and radial basis function centers of the radial base layer, and the weight of the output layer;
[0123] In one embodiment, the first optimization module 20 is specifically used for:
[0124] According to the difference between the reference energy consumption data and the predicted energy consumption data under different operating conditions, a genetic algorithm is used to optimize the number of nodes and the center of the radial basis function of the radial base layer and the weight of the output layer.
[0125] In one embodiment, the radial basis function is any one of a Gaussian function, a polynomial function or an inverse multiquadratic function.
[0126] Each module in the above parameter optimization device can be implemented in whole or in part by software, hardware, or a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to each module.
[0127] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Fig. 9 As shown. The computer device includes a processor, a memory and a network interface connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store sample vehicle data. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a parameter optimization method is implemented.
[0128] Those skilled in the art will understand that Fig. 9 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.
[0129] In one embodiment, a computer device is provided, including a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the following steps are implemented:
[0130] The sample vehicle data under different operating conditions are input into the radial basis function neural network to obtain the predicted energy consumption data under different operating conditions;
[0131] According to the reference energy consumption data and predicted energy consumption data under different operating conditions, the model parameters of the radial basis neural network are optimized by using a genetic algorithm; wherein the model parameters are associated with the parameters to be corrected of the vehicle's whole vehicle energy consumption mechanism model, and the whole vehicle energy consumption mechanism model is used to predict the vehicle's whole vehicle energy consumption data;
[0132] Based on the optimized model parameters, the parameters to be corrected are optimized.
[0133] In one embodiment, the vehicle energy consumption mechanism model is established in the following manner:
[0134] Obtain the vehicle's motor parameters, battery parameters and electrical accessory parameters; construct a drive motor mechanism model based on the motor parameters, construct a power battery mechanism model based on the battery parameters, and construct an electrical accessory system fitting model based on the electrical accessory parameters; based on the mechanism fusion model, establish the coupling relationship and constraints between the drive motor mechanism model, the power battery mechanism model and the electrical accessory system fitting model; based on the drive motor mechanism model, the power battery mechanism model, the electrical accessory system fitting model, the coupling relationship and constraints, establish a vehicle energy consumption mechanism model.
[0135] In one embodiment, sample vehicle data under different operating conditions are obtained by:
[0136] Acquire candidate vehicle data of the vehicle under different operating conditions; use the optimal Latin hypercube algorithm to sample the candidate vehicle data to obtain sample vehicle data of the vehicle under different operating conditions.
[0137] In one embodiment, the radial basis neural network includes an input layer, a radial basis layer and an output layer; when the processor executes the computer program, the following steps are also implemented:
[0138] For each operating condition, the sample vehicle data under the operating condition is input into the input layer for basic feature extraction to obtain basic feature data; the basic feature data is input into the radial base layer for nonlinear transformation processing to obtain nonlinear feature data; the nonlinear feature data is input into the output layer for weighted summation processing to obtain the predicted energy consumption data under the operating condition.
[0139] In one embodiment, the model parameters include the number of nodes and the center of the radial basis function of the radial base layer, and the weight of the output layer; when the processor executes the computer program, the following steps are also implemented:
[0140] According to the difference between the reference energy consumption data and the predicted energy consumption data under different operating conditions, a genetic algorithm is used to optimize the number of nodes in the radial base layer, the center of the radial basis function, and the weight of the output layer.
[0141] In one embodiment, the radial basis function is any one of a Gaussian function, a polynomial function, or an inverse polyquadratic function.
[0142] In one embodiment, a computer readable storage medium is provided, on which a computer program is stored, and when the computer program is executed by a processor, the following steps are implemented:
[0143] The sample vehicle data under different operating conditions are input into the radial basis function neural network to obtain the predicted energy consumption data under different operating conditions;
[0144] According to the reference energy consumption data and predicted energy consumption data under different operating conditions, the model parameters of the radial basis neural network are optimized by using a genetic algorithm; wherein the model parameters are associated with the parameters to be corrected of the vehicle's whole vehicle energy consumption mechanism model, and the whole vehicle energy consumption mechanism model is used to predict the vehicle's whole vehicle energy consumption data;
[0145] Based on the optimized model parameters, the parameters to be corrected are optimized.
[0146] In one embodiment, the vehicle energy consumption mechanism model is established in the following manner:
[0147] Obtain the vehicle's motor parameters, battery parameters and electrical accessory parameters; construct a drive motor mechanism model based on the motor parameters, construct a power battery mechanism model based on the battery parameters, and construct an electrical accessory system fitting model based on the electrical accessory parameters; based on the mechanism fusion model, establish the coupling relationship and constraints between the drive motor mechanism model, the power battery mechanism model and the electrical accessory system fitting model; based on the drive motor mechanism model, the power battery mechanism model, the electrical accessory system fitting model, the coupling relationship and constraints, establish a vehicle energy consumption mechanism model.
[0148] In one embodiment, sample vehicle data under different operating conditions are obtained by:
[0149] Acquire candidate vehicle data of the vehicle under different operating conditions; use the optimal Latin hypercube algorithm to sample the candidate vehicle data to obtain sample vehicle data of the vehicle under different operating conditions.
[0150] In one embodiment, the radial basis neural network includes an input layer, a radial basis layer and an output layer; when the computer program is executed by a processor, the following steps are also implemented:
[0151] For each operating condition, the sample vehicle data under the operating condition is input into the input layer for basic feature extraction to obtain basic feature data; the basic feature data is input into the radial base layer for nonlinear transformation processing to obtain nonlinear feature data; the nonlinear feature data is input into the output layer for weighted summation processing to obtain the predicted energy consumption data under the operating condition.
[0152] In one embodiment, the model parameters include the number of nodes and the center of the radial basis function of the radial base layer, and the weight of the output layer; when the computer program is executed by the processor, the following steps are also implemented:
[0153] According to the difference between the reference energy consumption data and the predicted energy consumption data under different operating conditions, a genetic algorithm is used to optimize the number of nodes in the radial base layer, the center of the radial basis function, and the weight of the output layer.
[0154] In one embodiment, the radial basis function is any one of a Gaussian function, a polynomial function, or an inverse polyquadratic function.
[0155] In one embodiment, a computer program product is provided, comprising a computer program, which, when executed by a processor, implements the following steps:
[0156] The sample vehicle data under different operating conditions are input into the radial basis function neural network to obtain the predicted energy consumption data under different operating conditions;
[0157] According to the reference energy consumption data and predicted energy consumption data under different operating conditions, the model parameters of the radial basis neural network are optimized by using a genetic algorithm; wherein the model parameters are associated with the parameters to be corrected of the vehicle's whole vehicle energy consumption mechanism model, and the whole vehicle energy consumption mechanism model is used to predict the vehicle's whole vehicle energy consumption data;
[0158] Based on the optimized model parameters, the parameters to be corrected are optimized.
[0159] In one embodiment, the vehicle energy consumption mechanism model is established in the following manner:
[0160] Obtain the vehicle's motor parameters, battery parameters and electrical accessory parameters; construct a drive motor mechanism model based on the motor parameters, construct a power battery mechanism model based on the battery parameters, and construct an electrical accessory system fitting model based on the electrical accessory parameters; based on the mechanism fusion model, establish the coupling relationship and constraints between the drive motor mechanism model, the power battery mechanism model and the electrical accessory system fitting model; based on the drive motor mechanism model, the power battery mechanism model, the electrical accessory system fitting model, the coupling relationship and constraints, establish a vehicle energy consumption mechanism model.
[0161] In one embodiment, sample vehicle data under different operating conditions are obtained by:
[0162] Acquire candidate vehicle data of the vehicle under different operating conditions; use the optimal Latin hypercube algorithm to sample the candidate vehicle data to obtain sample vehicle data of the vehicle under different operating conditions.
[0163] In one embodiment, the radial basis neural network includes an input layer, a radial basis layer and an output layer; when the computer program is executed by a processor, the following steps are also implemented:
[0164] For each operating condition, the sample vehicle data under the operating condition is input into the input layer for basic feature extraction to obtain basic feature data; the basic feature data is input into the radial base layer for nonlinear transformation processing to obtain nonlinear feature data; the nonlinear feature data is input into the output layer for weighted summation processing to obtain the predicted energy consumption data under the operating condition.
[0165] In one embodiment, the model parameters include the number of nodes and the center of the radial basis function of the radial base layer, and the weight of the output layer; when the computer program is executed by the processor, the following steps are also implemented:
[0166] According to the difference between the reference energy consumption data and the predicted energy consumption data under different operating conditions, a genetic algorithm is used to optimize the number of nodes in the radial base layer, the center of the radial basis function, and the weight of the output layer.
[0167] In one embodiment, the radial basis function is any one of a Gaussian function, a polynomial function, or an inverse polyquadratic function.
[0168] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to the memory, database or other medium used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetoresistive random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM). The database involved in each embodiment provided in this application may include at least one of a relational database and a non-relational database. Non-relational databases may include distributed databases based on blockchains, etc., but are not limited to this. The processor involved in each embodiment provided in this application may be a general-purpose processor, a central processing unit, a graphics processor, a digital signal processor, a programmable logic device, a data processing logic device based on quantum computing, etc., but are not limited to this.
[0169] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above 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.
[0170] 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 understood as limiting the scope of the present 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 present application shall be subject to the attached claims.
Claims
1. A parameter optimization method, characterized in that: The method comprises: The sample vehicle data under different operating conditions are input into the radial basis function neural network to obtain the predicted energy consumption data under different operating conditions; According to the reference energy consumption data and the predicted energy consumption data under different operating conditions, the model parameters of the radial basis neural network are optimized by using a genetic algorithm; wherein the model parameters are associated with the parameters to be corrected of the whole vehicle energy consumption mechanism model of the vehicle, and the whole vehicle energy consumption mechanism model is used to predict the whole vehicle energy consumption data of the vehicle; Based on the optimized model parameters, the parameters to be corrected are optimized.
2. The method according to claim 1, characterized in that The vehicle energy consumption mechanism model is established in the following way: Obtain the vehicle's motor parameters, battery parameters, and electrical accessory parameters; Constructing a drive motor mechanism model according to the motor parameters, constructing a power battery mechanism model according to the battery parameters, and constructing an electric accessory system fitting model according to the electric accessory parameters; Based on the mechanism fusion model, a coupling relationship and constraint conditions are established among the drive motor mechanism model, the power battery mechanism model and the electric accessory system fitting model; The whole vehicle energy consumption mechanism model is established based on the drive motor mechanism model, the power battery mechanism model, the electric accessory system fitting model, the coupling relationship and the constraint conditions.
3. The method according to claim 1, characterized in that: The sample vehicle data under the different operating conditions are obtained in the following manner: Acquiring candidate vehicle data of the vehicle under different operating conditions; The candidate vehicle data is sampled using an optimal Latin hypercube algorithm to obtain sample vehicle data of the vehicle under different operating conditions.
4. The method according to claim 1, characterized in that The radial basis neural network comprises an input layer, a radial basis layer and an output layer; The sample vehicle data under different operating conditions are input into the radial basis function neural network to obtain the predicted energy consumption data under different operating conditions, including: For each operating condition, inputting sample vehicle data under the operating condition into the input layer to extract basic features to obtain basic feature data; Inputting the basic characteristic data into the radial base layer for nonlinear transformation processing to obtain nonlinear characteristic data; The nonlinear characteristic data is input into the output layer for weighted summation processing to obtain the predicted energy consumption data under the operating condition.
5. The method according to claim 4, characterized in that The model parameters include the number of nodes and radial basis function centers of the radial base layer, and the weight of the output layer; The method of optimizing the model parameters of the radial basis function neural network by using a genetic algorithm according to the reference energy consumption data and the predicted energy consumption data under different operating conditions includes: According to the difference between the reference energy consumption data and the predicted energy consumption data under different operating conditions, a genetic algorithm is used to optimize the number of nodes and the center of the radial basis function of the radial base layer and the weight of the output layer.
6. The method according to claim 5, characterized in that The radial basis function is any one of a Gaussian function, a polynomial function or an inverse multiquadratic function.
7. A parameter optimization device, characterized in that: The device comprises: An acquisition module is used to input sample vehicle data under different operating conditions into a radial basis function neural network to obtain predicted energy consumption data under different operating conditions; A first optimization module is used to optimize the model parameters of the radial basis neural network using a genetic algorithm according to the reference energy consumption data and the predicted energy consumption data under different operating conditions; wherein the model parameters are associated with the parameters to be corrected of the whole vehicle energy consumption mechanism model of the vehicle, and the whole vehicle energy consumption mechanism model is used to predict the whole vehicle energy consumption data of the vehicle; The second optimization module is used to optimize the parameters to be corrected based on the optimized model parameters.
8. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 6 are implemented.
9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.
10. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 1 to 6 are implemented.