U-model-based adaptive control for a class of stochastic non-linear dynamic plants with unknown parameters

In this paper, an adaptive control algorithm within a U-model framework is developed for controlling a class of stochastic non-linear discrete-time models with unknown parameters. With the authors' previous justification, the control-oriented model not only represents a wide range of smooth (polynomial) non-linear dynamic plants (without using linearisation approximation at all), but also makes almost all linear control system design techniques directly applicable to non-linear dynamic plants (with a root solver bridging the linear design and calculation of controller output). A new recursive least squares algorithm is derived and its convergence is proved for the online estimation of time-varying parameters. For initial bench test, a pole placement controller for non-linear stochastic polynomial models is designed using the corresponding linear design technique. Accordingly a step by step procedure is listed to implement the adaptive control operation. A number of simulated case studies are conducted to illustrate the efficiency of the claimed insight and design procedure. Copyright © 2011 Inderscience Enterprises Ltd