Joint control of a walking robot

A system capable of controlling the leg operation of a mobile walking robot travelling over variable terrain must be capable of adaptive reaction in negotiating unpredictable irregularities in the working floor surface. This paper proposes a motion planning control architecture for a pneumatically powered mobile robot based on the analysis of leg co-ordination and leg trajectory given a demanded body velocity vector. Emphasis is placed on the control of the individual leg joints and a joint control system that uses both fuzzy logic and genetic algorithms as adaptive operators is proposed. The final solution is tested on an experimental rig and the results are compared with those obtained using conventional control alternatives.