Super-Twisting Second-Order Sliding Mode Control for Automated Drifting of Distributed Electric Vehicles 2020-01-0209

Studying drifting dynamics and control could extend the usable state-space beyond handling limits and maximize the potential safety benefits of autonomous vehicles. Distributed electric vehicles provide more possibilities for drifting control with better grip and larger maximum drift angle. Under the state of drifting, the distributed electric vehicle is a typical nonlinear over-actuated system with actuator redundancy, and the coupling of input vectors impedes the direct use of control algorithm of upper. This paper proposes a novel automated drifting controller for the distributed electric vehicle. First, the nonlinear over-actuated system, comprised of driving system, braking system and steering system, is formulated and transformed to a square system through proposed integrative recombination method of control channel, making general nonlinear control algorithms suitable for this system. On this basis, a super-twisting second-order sliding mode controller is designed to generate the desired virtual control input, pushing the vehicle to the performance and operating limits while keeping in the quasi-equilibrium state. Finally, an online control allocation is designed to translate the virtual control input to the longitudinal and lateral forces of each wheel. Optimization algorithm is used in the online allocation to guarantee stability performance. Co-simulations are conducted between two commercial software, i.e. CarSim and MATLAB/Simulink. The results have verified the effectiveness of proposed automated drifting controller for distributed electric vehicles.