State And Parameters Estimation For Distributed-electric-drive Vehicles

The distributed-drive electric vehicle has a simple structure,in which the advantages lie in the independent control of all-wheel-drive torque distribution with fast response and accuracy.It shows great potential in vehicle dynamics controllability and economic performance improvement.To elaborate the advantages of distributed-drive electric vehicles,it is needed to design a high-performance distributed-drive control system.Moreover,the precise values of vehicle state and parameters such as vehicle speed and mass are required.Therefore,an accurate estimation of vehicle state and parameters are necessary.This thesis analyzes and establishes two vehicle dynamics models in prior to parameter estimation and state estimation for the distributed-drive electric vehicles.The vehicle mass and road slope are estimated based on recursive least squares with variable forgetting factors,and the vehicle speed and the sideslip angle are estimated based on an unscented Kalman filtering.The drive motor speed and torque,longitudinal vehicle acceleration and longitudinal vehicle speed are used to observe the vehicle mass and road slope.Based on the longitudinal vehicle acceleration,lateral vehicle acceleration,yaw rate,and steering wheel angle as well as other data from lowcost sensors,the longitudinal speed,lateral speed,sideslip angle can thereby be observed.To establish Simulink-CarSim co-simulation platform,the vehicle model with its parameter setting was built in CarSim,and the state and parameter estimators were built in Simulink.The effectiveness of the proposed estimators in estimating the vehicle mass,longitudinal vehicle speed,and sideslip angle can be verified by co-simulation under typical operating conditions.The National Instruments c-RIO real-time controller and the software LabVIEW were used to build the distributed-drive electric vehicle rapid control prototype(RCP)test platform for testing the performance of the vehicle speed estimation.RCP tests were carried out under three working conditions including straight-line acceleration,turning and single lane change.The estimated vehicle speed in the test was compared with high-precision GPS data to validate the effectiveness of the vehicle speed estimation algorithm.