Optimization of a Robust Modified Smith Predictor Control Strategy for Integrating Processes with Dead Time

The presence of an integrator and dead time in physical processes reduces stability and robustness. It limits the response time of a system. Integrating plus dead time (IPDT) processes provide oscillatory and slow response if the parameters of a system are not tuned properly despite dead time compensators (DTCs) being used. To overcome these shortcomings, the Smith predictor based sliding-mode control (SP-SMC) strategy using the Jaya optimization technique for IPDT processes is proposed in this study. For the selected populations, the cost function and best controller parameters are evaluated. The proposed strategy is compared with the typical Smith predictor-based proportional, integral and derivative (SP-PID), and conventional SP-SMC design methods. To evaluate the performance, integrating first-order with dead time (IFODT) process models with different controllability relationships (CR) is considered. Robustness analysis of the controller is carried out in this study for 30% parametric uncertainties and bounded disturbances. The simulation tests on a laboratory process (level) control system reveal the supremacy of the Jaya optimization algorithm over prevalent control strategies. Compared to SP-PID and SP-SMC, the proposed design method shows an improvement of 33.07% and 12.58% in settling time and an improvement of 19.73% and 22.93% in rise time with 0% overshoot, respectively. The applied setup elicits better multi-level set point tracking and disturbance rejection capabilities with the step input. Besides, the proposed algorithm shows better closed-loop performance for numerical simulations of Models 1 and 2.