Parameter Optimization of Pure Electric Vehicle Power System Based on Genetic Algorithm

In this paper, the ADVISOR software was used to establish a complete vehicle model of an electric vehicle, and the model was verified by CYC_NEDC under European urban conditions to meet the requirements. The maximum power of the driving motor, the speed ratio of the transmission system and the capacity of the storage battery are taken as the optimization objectives to carry out multi-objective optimization. Connect the model built by genetic algorithm and ADVISOR, run the program to simulate the two together, and get the result of parameter optimization of dynamic system. Through the simulation analysis and comparison under CYC_NEDC cycle conditions, the maximum speed, maximum climb slope, acceleration time and other dynamic performance parameters of this electric vehicle are effectively improved after optimization.


Introduction
For pure electric vehicles, how to design the power system and optimize the parameters so that they can achieve the best power performance under different driving conditions, meet the maximum speed, acceleration and climbing requirements. At the same time, choose an effective way to optimize and match the corresponding performance parameters, so that under the premise of meeting the requirements, each performance index more. At present, for the optimization of power system parameters of pure electric vehicle, the transmission speed ratio is mainly chosen as the optimization design variable. However, only optimizing the transmission speed of pure electric vehicle has some limitations to improve the vehicle performance. Therefore, this article has carried on the improvement www.scholink.org/ojs/index.php/jetr Journal of Economics and Technology Research Vol. 1, No. 2, 2020 2 Published by SCHOLINK INC. in this aspect.

2.1Parameter Selection and Matching
This paper selects a pure electric vehicle for research, its basic parameters and performance indicators as shown in Table 1. According to the basic parameters of the whole vehicle, the selection of the battery and the motor and the calculation of the transmission ratio of the transmission system, the matching result of the power system parameters is as shown in Table 2.

Establishment and Verification of Vehicle Motion Module
According to the structure of Ev and the step-by-step transfer mode of power flow, the whole ev simulation model is established with ADVISOR software, as shown in Figure 1. The whole vehicle motion module is established by backward simulation after analyzing the vehicle motion equation and all the resistance during the vehicle running, as shown in Figure 2.  Table 3.    Table 3 with the requirements of the performance indexes in Table 1, it is concluded that the dynamic performance indexes of the electric vehicle model meet the design requirements under CYC cycle working condition, and the real-time vehicle speed and the circulation condition request the speed to follow the situation to be good.
Therefore, the calculated matching parameters of EV power system meet the requirements of power performance.

Establishment of Parameter Optimization Model of Power System Based on Genetic Algorithm
On the basis of the correct modeling, the parameters of the model are optimized and analyzed to realize the optimization of the dynamic system. In this paper, genetic algorithm is used to optimize the power system parameters of EV. Firstly, the parameter optimization model is established.

Selected Variable
In this paper, the maximum power Pmax of the drive motor, the transmission ratio I 0 and the battery capacity C are selected as optimization variables to ensure the optimal dynamic performance of the vehicle during the driving process, the selected variables are as follows:

Establish an Optimal Objective Function
The optimization objectives of power performance of pure electric vehicle mainly include the vehicle's maximum speed UAMAX, maximum climbing degree Imax, and minimum acceleration time T.
Compared with the first two indexes, the acceleration time can better reflect the comprehensive power performance of the vehicle. Therefore, the acceleration time is chosen as the optimization target to measure the dynamic performance:

Establish Constraints for Optimization
On the premise of considering the optimization goal synthetically, the vehicle should also meet the vehicle performance index (Table 2) and the driving requirements, therefore, the following constraints are established: 1) Meet the constraints of the maximum speed 2) The constraint condition of the maximum climbing degree is satisfied 3) The acceleration time condition is satisfied 5) It meets the restriction condition of vehicle normal running and adhesion performance

Analog Operation
Based on the above modeling, the ADVISOR simulation model is called by genetic algorithm, and the ADVISOR performance is calculated by assigning optimization variables, and the optimization objective function is calculated, optimization complete.
The working process of power system parameter optimization of pure electric vehicle is shown in

SET GENETIC ALGORITHM RUN parameters
According to the rules of genetic algorithm and the characteristics of optimization target, the parameters of optimization model are set as follows:

Table4. Optimization of Model Parameters
Model parameter Numerical value

Population size 30
Maximum Evolution Algebra 30 Cross probability 0.9 Mutation probability 0.9 After determining the basic parameters of the Genetic Algorithm, the optimization program of the Genetic Algorithm is written, which connects the genetic algorithm and the ADVISOR model.   Table 6 shows the comparison of power performance indexes before and after the optimization of power system parameters of EV. After the optimization of power system parameters of EV, the maximum speed increased from 124.6 km/h to 126.8 km/h, and the maximum speed increased by 1.77%. Figure 6 shows the speed comparison curve before and after optimization under CYC cycle conditions. The acceleration time before and after optimization was reduced from 14s to 13.7 S and the acceleration time was reduced by 2.15%. After optimization, the acceleration performance was improved. In the aspect of hill climbing, the maximum hill climbing degree of the optimized ev was 27.6%, which was 2.6% higher than that before the optimization.

Result Analysis
In this paper, the multi-objective optimization of the power system is carried out by considering the relationship among the components of the powertrain, using the rules of genetic algorithm and the requirements of the optimization objective. Using the established vehicle model to carry on the simulation analysis to the optimized power parameter, obtains this vehicle power performance to enhance obviously. In order to provide the basis for the vehicle design in the next step, the simulation running analysis of pure electric vehicle is carried out under the specific working condition, and then the optimum power parameter matching is determined.