Stable Adaptive Control of Non-Collocated Systems Using the Progressive Learning Method

Abstract

A novel approach to adaptive control of non-collocated robotic systems is presented, and its feasibility and effectiveness are verified through experiments using an ultra-high speed direct-drive robot having structural flexibility. This approach, called "progressive learning," allows the system to learn parameters recursively and progressively, starting with the ones associated with low frequencies and moving up to the ones with a full spectrum. Even though the system has non-collocated sensors and actuators and the relative order is three or higher, the learning process is guaranteed to converge by exciting the system with a particular series of reference inputs having an appropriate frequency spectrum. Experimental results are provided to demonstrate the effectiveness of this method. Implementation issues and practical considerations are addressed as well.