Comparative analysis of predictive control and performance for nonlinear interconnected tanks

Abstract

This article presents a comparative review of classical and advanced control strategies applied to interconnected tank systems with nonlinear behavior, focusing on model predictive control. A qualitative, descriptive–comparative approach was used, based on the analysis of the most recent scientific literature. Studies on PID, MPC, and NMPC controllers were reviewed, evaluating their effectiveness against disturbances, parametric variations, and operational constraints reported in the literature. Results indicate that PID provides simplicity and ease of implementation, with acceptable performance in linear systems, but is limited in nonlinear and multivariable dynamics. Model predictive control exhibited superior reference tracking, constraint handling, and robustness, particularly when nonlinear models and optimization techniques are integrated. Analysis of performance metrics reported in the literature (RMSE, ISE, ITSE) demonstrates that NMPC achieves significant error reduction, enhanced stability, and improved settling time. It is concluded that NMPC is an effective alternative for complex industrial systems, confirming and extending previous findings while providing guidelines for future research in real-world applications.