A time-optimal adaptive control system via adaptive switching hypersurface

This paper presents a technique for synthesizing a time-optimal adaptive control system. The technique consists of combining off-line memorization with simple on-line calculations to determine the control signal. The adaptive switching hypersurface consists of images of switching points of each plant in the extended state space (the present and past state variables). The approximation is obtained by linearizing the projection of the adaptive switching hypersurface in a subspace, and finally storing the parameters of the hyperplanes thus obtained in memory. The on-line calculation corresponds to a simple function generator. The function generator consists of the memory containing the values of parameters of the hyperplanes and a linear interpolator. The input of the function generator is the present and past state variables, and its final output is
+K
or
-K
. The technique is applied to the pitch axis adaptive regulator control of the F-101B aircraft in various flight conditions, and simulations of the system by the hybrid computer are obtained.