Eigenstructure assignment based Multiobjective Dynamic State feedback controller design for MIMO system using NSGA-II

The controller tuning of strongly interacting Multi-Input Multi-Output (MIMO) system is a challenging problem for the designer. This paper addresses the design of Eigenstructure Assignment (EA) based Multi-Objective Dynamic State Feedback Controller (MODSFC) for linear discrete MIMO system with significant interaction. The two important conflicting aspects considered in designing the controller are robust stability to parameter perturbation and smaller control gain to improve the transient response of the closed loop system. Multiobjective evolutionary algorithms were capable of solving challenging control problem with conflicting design objectives. Fast and Elitist Multi-Objective Genentic Algorithm (MOGA) known as Nondominated Sorting Genetic Algorithm-II(NSGA-II) is successfully applied for solving this multi-objective problem in this paper. The significance of EA based multiobjective controller design is that it is possible to incorporate the above two design aspects into eigenstructure as conflicting multiobjective criteria of minimum condition number of eigenvector and minimum norm of controller gain. Added value of this work is that the proposed controller uses complex valued chromosomes. Validation upon a benchmark three tank system illustrates the effectiveness of the proposed controller.