Research On Fault-Tolerant Control Of Intelligent Vehicles Against Faults Of Brake Actuators

Braking system is one of the most important components to ensure passengers'travelling safety and smooth traffic flow,regardless it is a traditional car or an intelligent vehicle.Brake-By-Wire(BBW)is the core technology of intelligent vehicle,and its continuous development will promote the arrival of the era of intelligent driving.For systems such as automobiles,spacecraft,nuclear power plants and chemical plants that require higher safety performance,it is necessary to design a control system that is both stable and reliable.In other words,this kind of system can not only maintain the stability of controlled system,but also tolerate potential faults of components.This thesis mainly focuses on the study of Fault-Tolerant Control(FTC)design based on the abrupt failure of automotive actuators,The research is conducted as follows:(1)Bayesian network is a machine learning method based on Probability Theory for the uncertain problems.This thesis applies relevant Bayesian Theorem to the research on the sudden failure of BBW' s actuators,and designs an innovative Bayesian Network Diagnosis Method.(2)The research of this thesis focuses on the planar motion of vehicle,choosing the longitudinal velocity,the lateral velocity and the yaw velocity of automotive mass center as the target state variables.Subsequently,a FTC algorithm is designed which can redistribute braking force of the rest of normally running wheels when BBW's actuators break down suddenly.Because of the coupling of different state variables and the movement of wheels in the driving process,the thesis builds a nonlinear dynamic model with seven degrees of freedom(DOF).(3)A hierarchical active FTC strategy which aims to deal with the failure of BBW's actuators consists of a high-level controller and a lower-level distributor.The resulting strategy allows the vehicle to track the desired planar motions.The upper part of the hierarchical strategy is a controller based on dynamic Sliding Mode Control(SMC),which is combined with a nonlinear dynamic model with seven DOF,and takes the ideal yaw velocity and sideslip angle as control referential targets.In the lower part,the distributor is based on control allocation method of the non-iterative reconfiguration.BBW's brake force efficiency shows the degree of utilization of frictional force between tire and road,whose weighted square sum is chosen as the optimal control objective of the cost function.The distributor can obtain the optimal reconfigurable redistribution scheme by solving the minimum of the cost function.(4)The feasibility of the hierarchical active FTC strategy has been analyzed in seven kinds of fault condition in brake system,including partial failure of single wheel,completed failure of single wheel,partial failure of two wheels on the same side,partial failure of two wheels on different sides,completed failure of double wheels on different sides,partial failure of three wheels,and partial failure of four wheels.The strategy has designed optimal control allocation scheme respectively according to the different failure modes(depending on the number of wheels failing in brake system)and different degrees of failure(partial or total).