Intelligent/adaptive control strategies for robot manipulators

Abstract

The objective of this thesis is to utilize the strategic advantages of intelligent/adaptive control and to develop such a paradigm to handle the complex problems in an expert domain. The two-level hierarchical structure of an intelligent/adaptive control scheme is considered by using robust adaptive controllers and fuzzy-logic rules. In the lower level, precise adaptive force/position controllers with fast convergence and robustness properties are considered where novel robust adaptive controllers with gain dynamics are introduced. In the upper level, the strategic rulebase is exploited to deal with uncertainties. Theoretical aspects of fuzzy-logic controllers are examined and analyzed in terms of stability issues. Guidelines for establishing a stable rulebase in the fuzzy feedback system are introduced. Finally, the quasi-static force control problem is selected as an example of integration of both levels including the manipulator dynamics and the uncertainty in the workspace.