Research On Identification Of Driver Characteristics And Application On Four Wheel Independent Drive And Steering Electric Vehicle

Electrification and intelligentization have become one of the hotspots in the research of automotive technology in the world.With the coming of auto electrification and intellectualization,the electric control system equipped on the car will be more and more rich.Driver's characteristics are considered in the research and development of electronic control technology can improve the intelligent level of the automobile effectively and improve the comfort of driving.Based on the National Natural Science Foundation of China “Study on dynamics control for four wheel independent drive and steering electrical vehicle considering driving characteristic(51675257)”,classification and identification of driver's characteristics and four wheel active steering control of four wheel independent driving and steering electric vehicle considering driver's steering characteristic are studied in this paper.First of all,research on the method of driver characteristics classification.The drivers with certain driving experience are selected to carry on driving simulator experiment.The data that can reflect the driver's characteristics are collected and the eigenvalues of experimental data are extracted.The K-means clustering algorithm is applied to cluster the eigenvalues of experimental data.According to the clustering results,the driver characteristics are divided into three categories: cautious,general and radical.Secondly,the identification method of driver's characteristics is studied based on Hidden Markov model.On the basis of classification of the eigenvalues of experimental data,the GM-HMM identification model of driver's acceleration and steering characteristics is established respectively based on HMM theory,and the identification model is verified.And the identification accuracy of the identification models is verified.Thirdly,the online identification method of driver characteristics is studied.The online identification model of driver characteristics is established based on BP neural network,and the identification accuracy of the model is verified.The online identification program of driver's characteristics is complied by MATLAB/Simulink/Stateflow software.Realizing on the driver's driving behavior information is collected,the eigenvalues are extracted,and the types of driver's acceleration and steering characteristics are identified online.To lay the foundation of four wheel active steering control considering driver's characteristics.Then,the four wheel active steering control method considering the driver's steering characteristic is studied.The reference model considering driver's steering characteristics is used at low speed.The reference models for drivers with different steering characteristics are established based on BP neural network at low speed;The Linear two degree of freedom model is used as reference model at high speed.The additional angle controller and angle distribution controller are designed.The additional angle controller is used to generate the corresponding additional angle according to the state of the car turning.The angle allocation controller is used to redistribute the four wheels according to the additional angle to,and realization of four wheel steering control.Finally,the experimental verification is carried out.The experimental conditions are designed,and drivers are selected to carry out experiments on driving simulator to verify the identification accuracy of online identification program for driver's characteristics,four wheel steering and control effect of four wheel steering control method.The verification results show that: the identification accuracy of online identification method of driver characteristics is high;the four wheel active steering control can realize the four wheel steering control which considers the driver's characteristic under the premise of ensuring the safety of the vehicle.