Self-Tuning Control

Abstract

In control literature there is a growing interest in adaptive control procedures. Adaptive control deals with the question of how to control (regulate) an ‘uncertain’ system, uncertain in the sense that its structure is unknown or changing over time and/or its model parameters are unknown. According to Kreiselmeier (1983) there are two approaches; the indirect one and the direct one, also called model reference approach. In the indirect approach the unknown model parameters are identified (estimated) and the current parameter estimates are used to update the control parameters. The term ‘indirect’ means that the control parameters are not adapted directly but indirectly through the model parameter estimates. The indirect approach is referred to as self-tuning or self-optimizing adaptive control and is subject of this study. In the model reference approach, the parameters of a suitably chosen controller structure are adapted directly in such a way that the output follows the output of a prespecified reference model. See for references e.g. Åström (1973), Goodwin (1981), Kendrick (1981), Saridis (1977), Landau (1974), Monopoli (1974) and Åström (1983).