Two Degree of Freedom Robot Adaptive Controller

Abstract

This paper combines the simple technique for adaptive robot control proposed by Bar-Kana and Guez and the exploratory lerning schedule proposed by Selinsky and Guez into a unified two degrees of freedom (DOF) robot controller, thereby improving the performance of both approaches. We first briefly summarize both approaches. We then propose our new 2 DOF controller and explain our motivation for its design. Next we present the main theorem and conclude with a two-joints rigid manipulator example.