Dynamics Modeling And Research On State Estimation For Multi-axle Distributed Drive Electric Vehicles

The distributed electric drive technology has become an important trend of the development of vehicles nowadays.With the development of key technologies of the distributed drive electric vehicle,they have more and more applications in multi-axle heavy vehicles.However,the driving conditions of multi-axle vehicles are complex and changeable,and the vehicles have high requirement for handling stability.And with the increase of the axles,the number of electric wheels has doubled and the freedoms of vehicle motion are increased accordingly,which brings some challenges to the dynamic analysis and control of multi-axle vehicles.Therefore,it is necessary to carry out the relevant research on the multi-axle distributed drive electric vehicle.Based on the research object of the multi-axle distributed drive electric vehicle and with the purpose of vehicle dynamics analysis and stability control,a vehicle dynamics model is established and the research of state estimation is carried out based on this model,which can estimate the specific states for vehicle stability control.Specific researches are implemented in several aspects as follows.Initially,a full vehicle dynamics model with 23 degrees of freedom,including the longitudinal,lateral,vertical,yaw,pitch and roll motion of the body,and rotational motion of 8 wheels,is established by MATLAB/Simulink for the research of the 8-wheeled distributed drive electric vehicle in this thesis.By simulation with MATLAB/SIMULINK and by comparison with the TruckSim software,the reliability and practicality of the dynamics model are verified.And then based on the common characteristic of the dynamics model of multi-axle distributed drive electric vehicles,a general simulation platform with dynamics models of four-axle,five-axle,six-axle and eight-axle distributed drive electric vehicle is built,which can provide theoretical guidance and model foundation for the further study of multi-axle vehicles.Furthermore,considering the accuracy and calculation of state estimation,the vehicle dynamics model is simplified and reestablished to be applicable to state estimation.Based on this model,an extended Kalman filter state observer and an unscented Kalman filter state observer are respectively proposed to estimate the longitudinal velocity,vehicle sideslip angle,roll angle,roll rate and yaw rate.The simulation results show the proposed unscented Kalman filter observer has higher accuracy in estimating the vehicle state than the extended Kalman filter observer.Additionally,considering that some vehicle parameters such as the total mass of the vehicle,yaw moment of inertia and the position of the center of mass are changeable or not easily obtained in fact,a vehicle parameters observer is added into the proposed unscented Kalman filter state observer to establish a Dual unscented Kalman filter observer,which can make joint estimation of vehicle states and parameters.Finally,taking the estimated states as control variables,a differential steering control system for engineering application is designed based on the corresponding control algorithm.Thereinto,the full vehicle dynamics model is used as the vehicle module of stability control and the vehicle state observer is added to estimate the required states for the differential steering control.The vehicle dynamics model and state observer established in this thesis can provide theoretical and engineering guidance for the research of multi-axle distributed drive electric vehicles.