Abstract
This paper presents new parameter-adjustment algorithms for adaptive systems and analyzes their properties.
Most of the conventional methods for designing adaptive systems require ideal, in other words, unrealistic design conditions that restrict the applications of the methods. Some of these conditions are due to the parameter-adjustment algorithms implemented in the systems, i. e., the present popular algorithms perform satisfactorily only under certain restrictive circumstances that cause the ideal design conditions. In order to apply the adaptive methods to a variety of existing plants, it is necessary to develop new parameter-adjustment algorithms that can perform satisfactorily under more realistic circumstances where the conventional ones cannot. This paper presents such three algorithms and analytically clarifies their attractive properties.
