Design, Simulation and Performance Analysis of Parametric Estimation Algorithms Applied to Model Reference Adaptive Control

Abstract

This article performed the design, simulation, and performance analysis between two parametric estimation algorithms, the Gradient Method (MG) and the Recursive Least Squares Method (MMQR), both applied to the Adaptive Control by Reference Model (CAMR) system.  The study of design techniques and control analysis, as well as the comparison of the methods presented here, enhance the ability of the designer to deal with practical problems effectively. The main contribution of the article was to apply and clarify the advantages of the methods presented.  Thus, the specific objectives were:  identify the plant to be controlled;  discretize the plant;  discretize the plan (iii) build the control law;    implement the identification algorithm;   and analyze  and analyze the simulated results.  From numerical simulations, we analyzed the performance of each algorithm and its respective advantages, advantages, and limitations. The MMQR has an excellent transient regime, but its computational cost was high. The MG  has the slowest accommodation time and has low computational demand when compared to the MMQR.  By taking into account the characteristics of each algorithm and having prior knowledge about the plant you want to control, such previous information helps you choose the algorithm, thus enhancing the better performance of the control system. 

DOI:https://doi.org/10.56238/devopinterscie-247