Task-space control algorithm for Stewart platform

A concise dynamic modeling and control on the task space for a six-leg Stewart platform is developed. Via the measurement on the legs the pose of end-effector can be computed with small incremental variation by the forward kinematics method. An explicit formulation including the significant physical terms for the model reference adaptive scheme is completed. A filter dynamics technique is used so that there is no need of acceleration measurement. The control algorithm is designed based upon the Lyapunov method in association with the index of parameter adaptation and tracking performance. Simulations indicate that not only the tracking error can be approached convergent to zero, but also the unknown parameters can be efficiently estimated.