Intelligent control for automotive manufacturing-rule based guided adaptation

Numerous industrial control problems, especially in automotive manufacturing, deal with control of MIMO nonlinear systems that are either static or have dynamics that can be ignored with respect to the sampling rate.<P> We expand the adaptive control algorithm by introducing a new control law that is derived from the LQR design. We analyze the convergence of the adaptive control algorithm under this control law and compare it to the control law based on the modified Levenberg-Marquardt (LM) algorithm. We also demonstrate how these two control laws can be dynamically adjusted to handle control constraints. Finally, we revise and simplify the algorithm for generation of the rule base of initial conditions