Development of Decoupling Scheme for Higher Order MIMO Process Based On Hybrid Genetic and Nelder-Mead Algorithm A. A. LASHEEN 1 ;A. M. EL-GARHY 2 ; E. M. SAAD 3 ; S.M.EID 4 1, 2, 3 Communication, Computer and Control Dept., Faculty of Eng. Helwan University, 4 Communication Dept. Faculty of Eng. Cairo University, Egypt Abstract:- Systems with more than one control loop are known as (MIMO), or multivariable systems. These systems are characterized by significant interactions between their inputs and outputs. Loop interactions can cause system instability. The problem of interaction can be alleviated by a proper choice of input-output pairings such that interactions will be minimized, this lead to input-output multi- variable system pairing, and development decoupling compensator unit. In this paper a generalized decoupling technique is proposed. The proposed technique uses relative gain array (RGA) to select proper pairing and hybrid genetic and nelder-mead algorithm (HGNMA) to estimate the optimal elements' values of steady state decoupling compensator unit. The proposed technique is applied on a two thermally coupled distillation column characterized with four Inputs-four outputs. This Proposed technique proves remarkable success in minimizing the interaction between the inputs and outputs except that output has been proper paired with. Keywords: Multiple input multiple output (MIMO), Relative gain array (RGA). Decoupling, Genetic Algorithm, (GA). Nelder-Mead Algorithm (NM). Hybrid Genetic Nelder-Mead Algorithm (HGNMA). 1. Introduction The problem of loop interactions and decoupling control of MIMO systems has been extensively studied [1-8], where multivariable processes are controlled. All of them dealing with the possible design procedures which can be summarized in; determine RGA of process, the appropriate selection of input-output pairs and design the appropriate decoupling compensation network. To design the state decoupling network it is necessary to estimate the elements values of the steady state decoupling matrix using detailed analytical techniques with very high mathematical burdens, specially for high order MIMO process. Different optimization techniques can be used to estimate the elements values of the steady state decoupling matrix. Genetic Algorithm (GA) and Particle Swarm Optimization (PSO) were used in [9]. In this paper, a hybrid method is used, the GA in conjunction with the nelder-mead simplex algorithm are combined together to have the better local and global optimization searching abilities simultaneously. HGNMA is used to estimate the optimal values of the elements of the steady state decoupling compensation matrix. The proposed HGNMA technique can be easily applied on any high order MIMO process. The paper is organized as follows: Section 2 reviews the decoupling concepts, Section 3 introduces the genetic and nelder-mead algorithm (HGNMA) technique, Section 4 explains the proposed decoupling scheme based on HGNMA technique, Section 5 describes the thermally coupled distillation columns process as a case study. The proposed technique is simulated and applied on the case study; the results are evaluated and given in Section 6. Finally, the conclusion is presented in Section 7. Proceedings of the 11th WSEAS International Conference on Automatic Control, Modelling and Simulation ISSN: 1790-5117 544 ISBN: 978-960-474-082-6