Road Friction Coefficient Estimation Of In-wheel-motor Electric Vehicle

In recent years,energy crisis and environmental problems that traditional fuel vehicles brought about are increasingly apparent.Besides,vehicle active safety problems get much more attention.In this background,in-wheel motor electric vehicle is considered as one of the effective solutions to these problems.Compared to traditional fuel vehicles,in-wheel-motor electric vehicle has a compact structure and high transmission efficiency.It has a good performance in terms of steering,driving,braking and controllability due to the independent adjustable performance of the vehicle driving,braking torque.At the same time,road friction coefficient estimation is one of the core problems of vehicle active safety control system.Therefore,it is of great theoretical and engineering application value to accurately estimate road friction coefficient on in-wheel motor electric vehicle.This paper focuses on the estimation of road friction coefficient on in-wheel motor electric vehicle.It analyzes vehicle dynamics and tire models and builds state observers based on differential model.Finally,it puts forward estimation methods of road friction coefficient under straight-driving condition and steering condition respectively.The details are as follows:First of all,this paper analyzes the vehicle dynamics,puts forward the tire vertical load correction method,gives wheel dynamics model and Dugoff tire model,and analyzes tire nonlinear characteristics as well as the general form of Dugoff tire model.On this basis,it compares this model with high-precision vehicle dynamics model veDYNA to analyze and verify the vehicle dynamics model and the accuracy and effectiveness of the tire model.On this basis,this paper presents an estimation method based on the friction rate and longitudinal slip rate curve,which can effectively solve the problem of road friction coefficient estimation under straight-driving condition.Kalman filter is designed to estimate the wheel longitudinal slip rate based on longitudinal slip rate differential model.Based on wheel dynamic model,the sliding mode observer is designed to estimate the tire longitudinal force.Combining with the front tire vertical load correction formula,the friction rate is estimated.After obtaining the wheel longitudinal slip rate and wheel friction rate,road friction coefficient can be obtained at the bases of friction rate and longitudinal slip rate calibration curve by looking up the table.In order to estimate road friction coefficient under steering condition,this paper presents a method based on BP neural network.The method proof process is given based on vehicle dynamics equation and Dugoff tire model formula.After theoretical analysis,training samples are obtained,and BP neural network structure is designed.After well training neural network,simulation experiments are designed to validate the effectiveness of BP neural networks.