Research On Experiment Platform Of By-wire Electric Vehicle With Four Wheels Independent Contorl And Its Integrated Chassis Control Strategy

Four-wheel-independent electric vehicle with x-by-wire is a new type of vehicle of whichall wheels can be driven, brake and steered independently. It inherits all the advantages of theelectric vehicle driven by in-wheel-motor, and its steering system be free from the traditionalmechanical steering mechanism. Compared with conventional vehicles, the electric vehicle hasmore controllable degrees of freedom, so the maneuverability be greatly improved bycompleting the special actions such as zero radius turning, oblique driving and crab. Through theintegrated design of the chassis control system, the steering, drive and braking system can becoordinated to ensure each tire of the vehicle having maximum adhesion margin, that enhancesthe vehicle stability greatly. Above all, Four-wheel-independence electric vehicle with x-by-wirerepresents the future direction of the development of electric vehicles, and it is one of the hotresearch fields of vehicles.This Ph.D. dissertation is based on the project of National Natural Science Foundation ofChina named―Research on Control methods and Key Technology for X-By-Wire‖(No.50775096) and―Research on bilateral control methods of vehicle Steer-by-wire System byJoystick‖(No.51105165). Based on the summary of the relevant research results, the papermakes intensively studies on the research of the experiment platform of four-wheel-independentelectric vehicle with x-by-wire and its integrated chassis control strategy, including the research of the platform’s framework and control systems, the estimation of the state information whichare used in the control and the integrated control strategy of the drive/braking/steering systems.The specific contents of each part are as follows:(1) The framework of the experiment platform and the vehicle control system is studiedbased on full electronic control scheme, and the application layer protocols of the vehicle CANcommunication network is designed.Firstly, the framework of the experiment platform is built according to the analysis of thefunctional requirements. In-wheel motors, electromagnetic brakes and torque motors are used asthe actuators of the drive/steering/braking systems respectively. All these subsystemsinterchanges information via CAN bus. Then based on the framework scheme, the vehiclecontrol system is designed by modular design method, including the sensor unit, the vehiclecontrol unit and the execution unit. During the design process, all these submodule not only havethe basic function, but also are considered the fault-tolerant control strategy to improve thereliability. Moreover, the vehicle CAN network architecture is built and the application layerprotocols is designed according to the SAEJ1939, which can enhance the scalability of thesystem.(2) Based on the characteristics of the four-wheel-independent electric vehicle withx-by-wire, the method of vehicle state estimation using Unscented Kalman filter theory isstudied.In order to solve the problem of lacking of vehicle status information in the integratedcontrol, the vehicle state estimation for four-wheel-independent electric vehicle with x-by-wireis studied. Firstly, the three degrees of freedom estimation model including longitudinal, lateraland yaw is set up. Combining this model with the calculation of longitudinal forces and lateraltire force based on the single-track model, the UKF state estimator is designed and verified inthe simulation environment. According to the results of the verification, the estimator isoptimized with HSRI tire model and the estimation of the total vehicle’s mass, which can reducethe influence of the lateral force estimation errors and the quality changes to the state estimationaccuracy. The results of the simulation show that the algorithm has good estimation accuracy andthe estimated value of the vehicle’s mass converge to the true value quickly. (3) Combining with the hierarchical control thought, the integrated control strategy for thevehicle drive/braking/steering systems is designed using model predictive control algorithm withconstraint.Combined with the hierarchical control thought, the integrated control strategy of thevehicle drive/braking/steering systems is designed, which have considerations of thecharacteristics of the electric vehicle four wheel longitudinal force and lateral forceindependently controllable. The control strategy is divided into the integrated control layer andthe constrained control allocation layer. Integrated control layer applying the model predictivecontrol algorithm with feedback correction features, optimizes the vehicle control force, that islongitudinal resultant force, lateral resultant force and additional yaw moment torque. It aims atkeeping track of the reference model longitudinal speed, lateral speed and yaw rate. The processof the optimization considered the constraint problems caused by the actuators and roadadhesion conditions. Control allocation layer set each wheel tire adhesion margin maximum asthe target to distribute the four wheel’s longitudinal driving force and wheel angle. In order tosolve the nonlinear constrained problem and simplify the calculation of the solution process,control allocation layer is divided into equality constrained optimization and inequalityconstrained optimization. The results of the simulation show that the control strategy caneffectively track the ideal goal; meanwhile, four tires have approximately equal adhesionmargin.(4) The functional test for the established experiment platform is carried out. The designedstate estimator and integrated control strategy are all verified on the platform.The conditions including straight-line acceleration, four wheel steering, zero radius turningand crab are chosen to verify the functionality and reliability of the experiment platform. Amongthem, the four wheel steering condition is used to comprehensively test the performance of thedriving, steering and the sensor system. The results show that all the subsystems work normally.Based on the platform, the proposed UKF state estimator is verified in straight-line accelerationand large lateral acceleration conditions. The results show that the estimated values were in agood agreement with the real data. Finally, the integrated control strategy is experimented underthe various conditions. The test results indicate that the strategy perfectly solve the problem of running deviation and can track the reference target greatly under the corning condition.The study of this paper makes the following innovations:(1) The experiment platform of four-wheel-independent electric vehicle with x-by-wire isbuilt. The platform with the advantage of x-by-wire technology, can realize multi-operationworking modes, such as four wheel steering, zero radius turning and crab. The vehicle controlsystem is designed by modular design method, and the system scalability and reliability areemphatically considered. All the submodule not only have the basic function, but also areconsidered the fault-tolerant control strategy to improve the reliability.(2) According to the characteristics that the four wheel speed and torque can be obtainedaccurately, the state estimator applying UKF theory is designed for four-wheel-independentelectric vehicle with x-by-wire. The estimator uses HSRI tire model as the calculation of lateraltire force, and introduces the vehicle mass into the estimation process. These ways improve theestimation accuracy effectively.(3) Combined with the hierarchical control thought, the integrated control strategy of thevehicle drive/braking/steering systems is designed, which have considerations of thecharacteristics of the electric vehicle four wheel longitudinal force and lateral forceindependently controllable. The strategy is divided into integrated control layer and controlallocation layer. In order to reduce the affection of the model error, the integrated control layerapplies model predictive control algorithm which enjoy feedback correction features, and theconstraints are introduced into its optimization process. Control allocation layer set each wheeltire adhesion margin maximum as the target to distribute the four wheel’s longitudinal drivingforce and wheel angle.