Evaluation Method For Dynamic Response Quality Of Powertrain In Mode Switching Process Of Hybrid Electric Vehicle

Hybrid electric vehicle(HEV)has become one of the most practical and feasible technical solution in the field of new energy vehicles due to its high dynamical performances,good economy performances and low pollution.HEV is equipped with two kinds of power sources: engine and motor.The torque fluctuation caused by the switching of power source may cause the vehicle to jitter,abrupt,etc.Commonly used automobile evaluation system is not suitable for this process.It is urgent to build a dynamic response quality evaluation system for the mode switching process of HEV.This paper focuses on the subjective and objective unified evaluation of the dynamic response quality of the powertrain during the mode switching process of HEV and the improvement methods.According to the characteristics of vehicle configuration,the mode switching process is divided into three categories,and three typical switching processes are selected for dynamic analysis.The corresponding dynamic analysis model of the powertrain in the mode switching process is established.The subjective evaluation grade is designed and the objective evaluation index system is constructed.The subjective and objective experimental data of the mode switching process are collected through experiment.The objective evaluation indexes are classified,screened and weighted by cluster analysis and principal component analysis.and the grade and objective index are integrated to obtain a sample library of the mode switching process of HEV.Based on neural network,neural network optimized by genetic algorithm and support vector machine,the subjective and objective unified quality prediction models are established respectively.The data of the sample library is used as the basis for model training and verification,and the accuracy of the three types of prediction models are compared.According to the prediction score of the prediction model,the structural parameters of the simulation model are optimized,and the technical measures to improve the dynamic response quality of the mode switching process of HEV are discussed.The results show that the established objective evaluation index system is suitable for the quality evaluation of the mode switching process of HEV,and the weights of the objective evaluation indexes corresponding to the three types of mode switching process are different.The subjective and objective unified prediction model based on support vector machine algorithm has the highest accuracy,which provides support for the optimization and improvement of the powertrain.