Optimizing control of uncertain plants under varying external inputs

The paper is concerned with on-line set-point optimization in a multilayer process control structure, under uncertainty in process models and estimates of slow-varying (drifting) uncontrollable external inputs (disturbances). The situation when the external inputs are varying slower than the controlled process dynamics but sufficiently fast to make the iterative steady-state set-point optimization not applicable is considered. The aim of the paper is to present an algorithm developed for tracking optimal process performance in the mentioned uncertain drifting process environment. The algorithm is a development of the dual ISOPE (Integrated System Optimization and Parameter Optimization) technique of iterative steady-state set-point optimization. After a formulation of the optimizing control task, the new algorithm is presented and discussed in the paper. Finally, simulation results are given showing that the proposed technique can be successful and indicating influence of the algorithm key design parameters on its performance.