Genetic Algorithm Based Linear Quadratic Regulator Design For Stabilization Of Active Magnetic Bearing System With Decoupler

Abstract

The Active Magnetic Bearing system (AMB) is a mechatronic tool which is used to suspend spinning parts of a machine so that they rotate without contact to the stationary part of the machine. AMBs are fundamentally unstable, highly nonlinear and non-minimum phase systems. As a result, this thesis provides research efforts focused on developing the best state-feedback control system for the AMB System's stability. To solve the laborious manual tuning of the weighting matrices in the design of the linear quadratic regulator the genetic algorithm (GA) is used. The system’s mathematical model has been developed and also the properties of the uncontrolled system have been analyzed. The model developed shows that the AMB system considered is a 2x2 MIMO system. Therefore, the interaction of the inputs with the outputs has been analyzed using relative gain array analysis. It is observed from the simulation results that the GA-tuned LQR has resulted better performance compared to the manually tuned LQR considering the main control loops (highly interacting I/O pairs). But the GA tuned LQR has successfully reduced the unwanted undershoot resulted by the inputs in the outputs as a result of slight coupling. From this, the proposed controller for the loosely interacting input-output pairs the GA-tuned LQR has managed to reduce the undershoot by 66.54%. This demonstrates that the GA-tuned significantly lessened the undesirable effect of the inputs on the outputs