Adaptive gait generation with Six-legged Walking Chair Robot using the hierarchical decentralized system

This paper proposes the method that a hexapod robot is traversed irregular terrain such as the stairs or slopes using the hierarchical decentralized system. A higher subsystem, analizes the stability while walking and instructs lower subsystems, is made to participate in the decentralized system of which a lower subsystem that one leg is controlled object. So the robot was achieved walking that adjusted to an environmental change in cooperated operation between subsystems. A conventional "Zero Moment Point (ZMP)" is a useful criterion of stability, but it does not evaluate the degree of stability. Therefore this paper proposes a "Acceleration Stability Margin". Amount of acceleration that ZMP is necessary for limit in stability area to move. Then Six-legged Walking Chair Robot achieved walking on irregular terrain that improved a movability while keeping the lowest stability margin by evaluating Acceleration Stability Margin at each environmental change.