Optimal detuning of multivariable proportional integral controller based on data-driven approach for an activated sludge process

The study deals with designing a multivariable centralized proportional-integral (PI) controller for an activated sludge process (ASP) by adopting a new tuning technique. To overcome the tedious task of manual tuning by a classical method that was based on Ziegler-Nichols tuning, a novel approach is developed by formulating the optimization problem and considering it along with the theoretical results to calculate the controller parameters. The proposed method is then compared with the reported method to demonstrate the advantages of the proposed approach. A nonlinear ASP model is selected to evaluate the performance of the designed controller. The simulation carried through the proposed optimization technique exhibits a significant improvement in closed-loop performance for tuning the centralized controller. It is observed that the proposed method minimizes the effort of tuning the controller significantly. The proposed controller achieves better load disturbance response and lower overshoot than the reported method. The proposed method could improve integral absolute error (IAE) and integral time absolute error (ITAE) performance indices, ranging between 18.69% to 53.80% and 16.42% to 52.97% for a step change in yr1$$ y{r}_1 $$ and yr2$$ y{r}_2 $$, respectively, which are better compared to the reported method. The study also assesses the control system's robustness for their uncertainties with the time constant and time delay.