Steering control of electric tracked vehicle based on second-order disturbance observer and multiobjective optimization

Abstract

Autonomous and remote-controlled tracked vehicles are freeing humans from exhausting off-road maneuvers. Advanced motion control is a necessity to achieve high mobility and enhanced safety with less dependence on operator skill. Tracked vehicles may slide laterally or roll over under large centrifugal forces. Precise yaw motion and sideslip prevention are required for steering controller development. A switching control architecture is proposed for the underactuated tracked vehicle in this study. Two control laws for yaw rate tracking and anti-sideslip are proposed respectively based on second-order disturbance observers (DO-2s) with a given bandwidth. The controller is optimized for the two objectives using Nash bargaining method. The proposed steering controller is verified on a small electric track vehicle. Under large disturbance, the optimized DO-2-based controller prevents potential sideslip and reduces the yaw rate tracking error by 42.6% compared with LADRC. The chattering induced by switching is moderate because the estimated disturbances are smoothly switched.