Pso tuning for a centralized dead time compensator applied to tito processes

In this paper, a dead time compensator (DTC) based on a centralized PI controller with Smith Predictor is designed. The proposed approach is applied to a two-inputs and two-outputs (TITO) process. The Davison Method is used to get initial tuning parameters for the DTC and then Particle Swarm Optimization (PSO) is utilized for getting the fine-tuning of parameters. The proposed controller is implemented on the Temperature Control Lab® (TCLab), it is modeled as a First Order Plus Dead Time (FOPDT) model using reaction curve method. Reference changes and uncertainty tests are performed. The DTC controller is compared against a centralized PI controller using performance indexes based on Integral Square Error (ISE) and Total Variation (TV).