Performance Indixes, Predictive Control, Level System, Recursive Least-Squares, Monte Carlo Method
Abstract
This article presents a comparative study of performance indexes between two predictive control techniques in a single-tank level system with first-order plus dead-time characteristics. The first technique involves control using the Predictive Minimum Variance Control method, which adjusts its control based on two principal tuning parameters, while the second technique, Dynamic Matrix Control, relies on tree tuning parameters. For the mathematical modeling of the studied plant, dynamic system identification is employed using Recursive Least Squares with the ARIMAX model, and the validation of the estimated model is conducted through the Monte Carlo method.
