Comparison of the self-tuning on-off controller with the conventional switching controllers

A self-tuning on-off control algorithm concept is presented. The controller contains a predictor, which is used to estimate the future system output sequences taking into account the whole set of future on-off input sequences over a specified number of time intervals. For every time interval the optimal input is selected by minimizing a suitable cost function. The prediction is based on a parametric system model, whose parameters are estimated. some simulation studies are included to assess the relative performance characteristics of the proposed algorithms. The performance of the self-tuning on-off controller is compared with that of conventional switching controllers.