Supervision based on partial predictors for a multimodel generalised predictive control: experimental validation on a semi-batch reactor

The multimodel approach has become a major research topic during the last few decades and unlike many other advanced techniques, it has also been successfully applied in industry. This paper describes the application of a multimodel generalised predictive control (MGPC) based on a commutation algorithm to the problem of a semi-batch reactor control. The commutation is controlled by a set of partial predictors which constitute a supervisor. The experimental results show that the multimodel generalised predictive control based on supervisor leads to good performances in terms of precision and tracking, relatively to the case where the direct adaptive generalised predictive control (DAGPC), is applied.