Control of a redundantly actuated power line inspection robot based on a singular perturbation model

Abstract

This paper focuses on the controller design of the walking system of a power transmission line inspection robot that is redundantly actuated by two wheels. After an analysis the kinematics and dynamic model of the driven system, the controller is designed as a master-slave model to avoid the ldquoconflictrdquo between the two wheels. To effectively damp the elastic vibration caused by the flexible manipulator, the control system was decomposed into two subsystems, a slowly-changing subsystem and a fast-changing subsystem based on a singular perturbation model. A hybrid control method, consisting of a RBF adaptive PD control which was applied to slow-changing subsystem and optimal control which was applied to fast-changing subsystem was presented. The RBF adaptive algorithm can estimate the unknown resistance on-line and keep the controller stable. Simulation results prove this method to be reasonable and effective.