Road Tire Friction Coefficient Estimation For Four Wheel Drive Electric Vehicle Based On Moving Horizon Estimation

Four wheel drive(4WD)Electric Vehicles(EVs),with simplified powertrain,visual control mode and intelligent operation,have become a hot topic in recent years.This paper mainly study the estimation of the adhesion coefficient of 4WD EV.Compared with the adhesion coefficient estimation system of traditional vehicle,the torque and speed information of 4WD EV are easy to obtain,so the estimation strategy is more comprehensive.Firstly,the simulation model of high precision 4WD EV is constructed in AMESim,and the validity of the model is verified by off-line simulation through a variety of typical conditions.The longitudinal velocity and lateral velocity of the vehicle are estimated by the three-degree-freedom model.Through the selection of formulas,the effective slip rate and side slip angle calculation formulas are selected by a variety of simulation tests.The motor's own physical parameters and structural materials make its driving and braking capacity are limited,so the actual output torque should not exceed the maximum output torque,and the maximum value of the output torque varies with the motor speed and the motor voltage,so the time-varying output torque maximum constraint information should be considered.The surface adhesion coefficient changes from 0 to 1,so the hard constraint information of the adhesion coefficient should be taken into account,so we can choose the approximate moving horizon estimation which can use the rich data and restrain the information and prevent the data explosion.The first method:the longitudinal force estimation is completed by using the motor torque and wheel speed information.The lateral force is estimated by the ratio of the longitudinal force and the lateral force with the HSRI tire model,and then the adhesion coefficient is estimated.The second method:through the transformation of the HSRI tire formula,the implicit formula of the adhesion coefficient is explicit,which makes the solution of the adhesion coefficient easier and simpler.According to the drawbacks of the two estimation methods,the third estimator is designed by combining the two methods,obtaining the adhesion coefficient data points above the mean value of the first method,these data points' adhesion coefficient in the second method are collected as the estimation result.Finally,the validity of the estimators are verified by the AMESim and Simulink joint simulation experiments,respectively,under the conditions of high adhesion coefficient surface,low adhesion coefficient surface and varying adhesion coefficient surface.The main content and innovation of this paper:1.This article makes full use of 4WD EVs' advantages: the driving torque,braking torque,wheel speeds are obtained essier.The first method uses the moving horizon estimation to estimate the wheel longitudinal force,taking into account the constraint information of the wheel torque,and then use the HSRI tire model to estimate the lateral force of the tire,and then estimate the adhesion coefficient with the constraint of the adhesion coefficient;In the second method,the implicit formula with the adhesion coefficient is transformed into the explicit form by using the deformation of the HSRI tire formula,so as to facilitate the calculation of the Jacobi state equation and simplify the estimation method.The third method is the combination of the two methods,the current adhesion coefficient information is recorded when the road is fully utilized under the corresponding conditions.2.The moving horizon estimation algorithm is used to estimate the adhesion coefficient,which can fully consider the constraint situation of the estimated physical quantity under the actual physical condition,so that the result is more in line with the objective actual change.And the method is not based on the empirical initial value,without giving accurate initial value of the adhesion coefficient.3.The high-precision 4WD EV model is built in the advanced simulation software AMESim,and the model is verified through the off-line simulation experiments in a variety of typical conditions.Finally,the validity of the estimator is verified by AMESim and Simulink joint simulation platform,at the high adhesion coefficient road surface,low adhesion coefficient road surface and varying adhesion coefficient road surface conditions.