A trajectory tracking control during autonomous docking process for single-axle reconfigurable UGV modular unit using sliding mode prediction control

Abstract

The reconfigurable modular unit (RMU) with single-axle dual-wheel configuration is the foundational component of the reconfigurable unmanned ground vehicle (RUGV), providing enhanced adaptability through independently controlled steering and driving capabilities. Achieving high-precision trajectory tracking during the autonomous docking process of RMUs is crucial to ensuring the scalability and flexibility of RUGV. This paper presents a trajectory tracking control strategy based on sliding mode predictive control (SMPC), specifically designed for RMU docking. To account for the unique structural characteristics and motion patterns of RMUs, this paper introduces generalized control variables, specifically the theoretical slip and radius angles, to describe the motion behavior. These variables are the foundation for developing a kinematic model suitable for control design. A dynamic simulation platform is developed to analyze the steady-state steering characteristics under different combinations of steering modes. The proposed SMPC controller adopts a prediction model in the Frenet coordinate system and incorporates a differential controller to compensate for tracking errors. Simulation results validate the effectiveness of the proposed approach, showing strong robustness under varying docking speeds and superior tracking accuracy compared with traditional methods. In particular, among the steering mode combinations evaluated, the differential steering + Ackermann steering mode delivers the highest tracking accuracy, effectively harnessing the structural advantages of the RMU.