Fuzzy Dynamic Modeling for Walking Modular Robot Control

The paper presents new concepts and approaches of multi-stage fuzzy method of walking robots using resolved acceleration control. Several compliant control methods were analyzed in order to obtain high performances in robot trajectory control, which generates position and force parameters for multi-stage fuzzy control, some of which include a dynamic model in loop control and some of which doesn’t include dynamic model. For real-time control of robot stability the mathematical modeling of the center of gravity position was realized, in order to allow control of the walking robot when moving on terrains with complicated configuration, and relations which are necessary for robot position coordinates of the center of gravity were established. The studies presented have demonstrated the possibility of implementing force control by resolved acceleration, where dynamics and kinematics stability are simultaneously achieved in rigid environments. The obtained results lead to a smooth transition, in walking robots movement, without discontinuities, from controlling in position to controlling in force and position. Key-Words: fuzzy robot control, walking robot modeling, hybrid position-force control, robot dynamic control.