Design of a Non Fragile Optimal Control Based Realization in Comparison between LQR and Observer Based Controller for Magnetic Levitation System

Abstract

In this study, we examine the magnet-based rail steering technology known as Magnetic Levitation. The primary goal is to determine which of two types of controllers, Linear Quadratic Regulators (LQR) and an Observer-based controller, can correctly suspend and push a train down a guide track composed of magnets. A state space model of a Magnetic Levitation system is developed so that it can carry out the required operation. MATLAB/SIMULINK is used to model the system's reaction. Unexpected behavior in the open loop demonstrated the instability of the developed model. Closed-loop analysis is performed using the results from the Linear Quadratic Regulator (LQR) and the Observer based Controller. For a variety of magnetic tracks, both controllers performed well. Compared to the Linear Quadratic Regulator (LQR) controller, the reaction time of the observer-based controller is much shorter. Observer based controller and Linear Quadratic Regulator (LQR) models are developed in Simulink. Furthermore, we examine several realization strategies for least fragility in controller implementation, including minimal realization, balanced realization, modal realization, and observer canonical realization. An ideal non- fragile controller design has been developed after a thorough analysis of the discrepancies between the various realization controllers in terms of rounding off error or truncation error. The computer-generated model was subjected to a variety of perturbations. Both open and closed loops are used to examine the data. The train was effectively suspended and pushed along the track, as shown by the closed-loop reaction.