Research On Braking Force Distribution Strategy For Electric Vehicle Integrate With Braking Intention Prediction

Since the problems of energy consumption and environmental pollution are increasingly stringent,electric vehicle(EV)becomes a research focal point since it is renewable and clean.However,due to the limitation of cruising range,their large-scale promotion and application are restricted.Braking energy recovery can significantly improve the cruising range of EV.Since the braking process is conducted by the driver,the integration of braking intention is of great significance to the study of braking force distribution strategy of EV.This study takes electro-hydraulic compound braking system of a front-drive pure EV as the research subject.The vehicle model is built based on Matlab/Simulink and the driving data of the vehicle under the real driving conditions is collected as the database.Relief F and RRelief F algorithm are used for data preprocessing.When predicting the braking intention,firstly,the master cylinder pressure representing the braking intensity is taken as an input based on the fuzzy C-means(FCM)algorithm,and the braking intention is divided into four categories: slight braking,medium braking,strong braking,and emergency braking.Then,based on the random forest(RF),with the characteristic parameters selected by Relief F as input and FCM clustering results as the target,the prediction model is trained for online prediction of braking intention.Since the braking intensity is naturally continuous and closely related to road type and driver driving habits,this study selects a nonlinear autoregressive with external input neural network(NARX)with feedback and memory functions to predict the braking intensity.With the characteristic parameters selected by RRelief F as the input and the pressure of the main cylinder as the target,the prediction model is trained based on NARX to predict the braking intensity online.Finally,a new braking force distribution strategy is proposed: in slight braking scenario,the braking energy is not recovered in consideration of battery life;in medium braking and strong braking scenarios,ensure safety while maximizing the energy recovery in consideration of the difference between road adhesion coefficient and braking intensity;in emergency braking scenario,an optimal slip rate controller based on model predictive control(MPC)is designed to minimize the braking distance and improve the braking safety of the vehicle.Simulation results show that the prediction accuracy of braking intention and braking intensity are closely related to the prediction period.When the prediction period is less than 0.5s,the prediction accuracy is very high.The adoption of a new braking force distribution strategy can significantly reduce the charging times of the battery,and the impact on energy recovery is minimal.In case of emergency braking,compared with ABS based on MPC control,the braking distance can be reduced.