Path following fault-tolerant control of distributed drive autonomous unmanned vehicle via adaptive terminal sliding mode

The distributed drive autonomous unmanned vehicle (DDAUV) with unknown fault input of steering system will seriously deviate from the desired path and affect driving safety. This paper proposes a fault estimation and fault-tolerant control (FTC) algorithm, which can improve the comprehensive performance of lateral stability and path following through the torque redundancy characteristics of steering motor and in-wheel motors (IWMs). Firstly, Bayesian algorithm is applied for fault estimation with the prior estimation based on model and posterior observation by sensors. Once steering system fault occurs, the path following model will be reconstructed, and the linear quadratic regulator (LQR) algorithm can dynamically adjust desired front wheel steering angle for path following compensation. Besides, the torque vectoring control of IWMs is applied to further compensate considering the saturation of steering motor. Adaptive terminal sliding mode (ATSM) algorithm is employed to calculate desired additional yaw-moment where the weight matrix of ATSM is adjusted dynamically based on cooperative game theory. Furthermore, an oriented torque distribution method is designed to generate required additional yaw-moment. Finally, the CarSim-MATLAB/Simulink co-simulation and HIL experiment are carried out to test the effectiveness. The results indicate that the proposed FTC algorithm can ensure the lateral stability and improve the path following accuracy with steering system fault.