A Hybrid Control Framework for Chemical Processes with Long Time Delay: Theory and Experiments

This paper proposes a hybrid control framework based on internal model concepts, sliding mode control methodology, and fractional-order calculus theory. As a result, a modified Smith predictor (SP) is proposed for nonlinear systems with significant delays. The particular predictive approach enhances the sliding mode control (SMC) controller's transient responses for dead-time processes, and the SMC gives the predictive structure robustness for model mismatches by combining the previous methods with fractional order concepts; the result is a dynamical sliding mode controller. A numerical example is considered to evaluate the performance of the proposed approach, where a step change, external disturbance, and parametric uncertainty test are performed. A real application in the TCLab Arduino kit is presented; the proposed method presented good performance with a little amount of chattering, and in the disturbance rejection case, the overshoot increased with an aggressive response; in both cases, better tuning parameters can improve the process response and the controller action.