Adaptive system control

This article presents an approach to adaptive control based on a critical reading of the literature in this field. It consists in combining a predictive control method with a robust adaptation algorithm under an appropriate supervision logic. The choice of predictive control is motivated by its generality and ease of implementation. The robustness of the parametric adaptation algorithm is particularly required to preserve the performance of the adaptive control system in the presence of measurement noise, load disturbances and neglected and/or time-varying dynamics. Supervision makes it possible to apply the principle of certainty equivalence with the caution required to ensure the integrity of the adaptive control system, including the performance of the underlying control system.

A set of simulation results is presented around a famous example to illustrate the problems of adaptive control systems, in this case parametric adaptation. Two applications are presented to demonstrate the applicability of adaptive control techniques for self-tuning and controller calibration. They concern two pilot processes: a chemical reactor and a flexible transmission system.