<title>New adaptable reference model adaptive control for slewing control of a flexible beam with an unknown tip-load</title>

Abstract

Conventional model reference adaptive control (MRAC) requires perfect model matching condition as a prerequisite check for a successful controller design. Although such condition may be satisfied easily for some mechanical systems, it is extremely difficult for adaptive slewing control of a flexible beam with an unknown tip payload. The reason being, firstly, there must be as many actuators as the number of assumed vibration modes to be placed along the beam to provide bending moment. Secondly, if the first condition is satisfied, it would still result in very high feedback gains even for a very small variation of tip payload mass. Therefore the conventional MRAC scheme is physically unapplicable for such application. This paper addresses this issue and provides a solution to the general problem where the actuation power is limited such that the conventional MRAC scheme would not work. A new scheme called Adaptable Reference Model Adaptive Control (ARMAC) differs from the conventional MRAC in that the reference model of the ARMAC is not fixed. The ARMAC inheres infinitely many model gain sets which depend on the estimation of tip payload mass. As the controller/estimator is searching for the true tip payload mass, the working reference model is also trying to fit itself into the best reference model possible. The ARMAC employs the `steepest descent' method which the general MRAC uses to search for the true tip payload mass without any explicit parameter estimation process. This way the ARMAC avoids the stringent perfect model matching requirement while preserving the direct adaptation nature of MRAC successfully.© (1995) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.