Controller recon"guration for non-linear systems

This paper outlines an algorithm for controller recon"guration for non-linear systems, based on a combination of a multiple model estimator and a generalized predictive controller. A set of models is constructed, each corresponding to a di!erent operating condition of the system. The interacting multiple model estimator, extended for systems with o!set, is utilized to yield a reconstruction of the state of the non-linear system and mode probabilities of the di!erent models in the model set. Based on this information, a standard cost function in predictive control is optimized under the assumption that the mode probabilities are constant over the maximum costing horizon. The algorithm is illustrated for two di!erent case studies * one with a linear model of one joint of a space robot manipulator, subjected to failures, and one with a non-linear model of the inverted pendulum on a cart. ( 2000 Elsevier Science ‚td. All rights reserved.