Design of adaptive PID controllers based on adaptive Smith predictor for ultra-local model control

In this paper, an ultra-local model control approach based on adaptive Smith predictor is proposed. The design of adaptive PID controller takes into account the estimation of variable time delay which is compensated by the addition of an adaptive Smith predictor. The purpose of this paper is to solve the online estimation problem of time delay thanks to the proposed identification method of ultra-local model parameters. A performance comparison between the proposed control approach and the Smith predictor control with classical PID is carried out. The numerical simulation results of the thermal process study with severe constraints and operating conditions show the superiority of the adaptive PID controller. The robustness with respect to noises, disturbances and system parameter uncertainties of control approaches are highlighted.