MODEL-BASED FEEDFORWARD PRECOMPENSATION AND VS-TYPE ROBUST NONLINEAR POSTCOMPENSATION FOR UNCERTAIN ROBOTIC SYSTEMS WITH WITHOUT KNOWLEDGE OF UNCERTAINTY BOUNDS .2./

In this paper, the robust nonlinear controller for an uncertain robot system is developed and characterized with a unified method. Based on deterministic approach, the control structure consists of two parts : In the first part, the primary control law is synthesized to precompen sate for the nominal system; and in the second part the adaptive versi on of robust controllers are utilized to postcompensate for the system uncertainties. The uncertainties assumed in this papar are bounded by higher-order polynomials in the Euclidean norms of system states with out knowledge of bounding coefficients. Using the Lyapunov stability t heory, we can guarantee that all possible responses of the closed-loop system are at least uniformly and ultimately bounded. The tracking pr operties of the control algorithms are verified through numerical simu lations, and the results show that the proposed controllers are proven to be robust enough for any higher-order system uncertainty.