Design of QFT Controller and Pre filter Algorithms For MIMO systems

Abstract

Control theory is a design synthesis problem wherein system dynamics are controlled to newlineget a desired output under system uncertainty. The robust control design approaches have newlineseen limitations due to inexact knowledge of system dynamics, system non-linearity, newlineinaccurate modelling of input-output disturbances, ignorance towards stochastic dynamics newlineetc. The synthesis and automation of control systems is of paramount importance and it is newlineseen as a practical difficulty in the design of control systems. Synthesis is concerned with newlineachieving a stable controller in order to make the closed-loop system meet a set of criteria newlinespecified for performance under plant uncertainty and disturbance. The control objective newlineis achieved by feedback, which is necessary for disturbance rejection and zero steady newlinestate error. It is essential for the controller to assure quality control over time, but plant newlinemodel imprecision, parameter variation; non-linearity etc. impacts this adversely. newlinePractical systems demonstrate diverse performances which are different from the newlinetheoretical model used during the design phase. Robust control design approaches such as newlineProportional-Integral-Derivative (PID) control, Linear Quadratic Regulator (LQR), newlineLinear Quadratic Gaussian (LQG), Hand#8734; Control, Fuzzy Control etc. along with newlineoptimization techniques such as Particle Swarm Optimization (PSO), Ant Colony newlineOptimization (ACO), Lion Optimization Algorithm (LOA), Genetic Algorithm (GA), newlineSimulated Annealing (SA) etc. are implemented with the quantitative feedback technique newline(QFT), but, all these methods do require generation of plant templates and bounds which newlineultimately puts limitations on novice designer s know-how and adds to complexity. newlineThe design engineers in the control system domain sees multi-input multi-output (MIMO) newlineplant having uncertainty as a challenging problem in real-time applications. The newlinefrequency domain method, because it offers a ready means of analyzing information for newlinemulti input multi output systems with uncertainty comes