Control Methods Study of Rail-Mounted W-Beam Guardrail Inspection Robot

To address the limitations of traditional manual highway guardrail inspections, this paper proposes an obstacle-crossing and collaborative tracking control method for a rail-mounted robot. Static and dynamic analyses verify the robot's structural reliability and driving feasibility. Based on the leader-follower model, a triangular collaborative tracking model is developed, and a linear time-varying model predictive controll (LTV-MPC) is designed to achieve smooth and precise collaborative control. For obstacle crossing, an acceleration reference model and a gradient-based adaptive law are proposed, leading to a model reference adaptive controll (MRAC) that effectively suppresses vibrations and ensures synchronous control. Simulation results show that the MPC achieves a 0.415% overshoot and a 0.344 m steady-state accuracy, while also reducing the intensity of speed fluctuations by 35%. The MRAC ensures smooth obstacle-crossing speeds and adaptive strategy switching, validating the reliability and practicality of the rail-mounted robot under complex working conditions.