PT Unknown
AU Wei, C
TI Controller Design and Optimization for Rotor System Supported by Active Magnetic Bearings
PD 12
PY 2015
LA en
AB Active Magnetic Bearings (AMBs) have been receiving increased attention in industry because of the advantages (contact-free, oil-free, etc.,) that they display in comparison with conventional bearings. They are used extensively in rotor system applications, especially in conditions where conventional bearing systems fail. Most AMBs are controlled by Proportional-Integral-Derivative (PID)-controllers. Controller design for AMB systems by means of hand tuning is time-consuming and requires expert knowledge. In order to avoid this situation and reduce the effort to tune the controller, multi-objective optimization with genetic algorithm is introduced to design and optimize the AMB controllers. In the optimization, criteria both in time and frequency domain are considered. A hierarchical fitness function evaluation procedure is used to accelerate the optimization process and to increase the probability of convergence. This evaluation procedure guides the optimizer
to locate the small feasible region resulting mainly from the  requirement for stability of control system. Another strategy to reduce the number of optimization parameters is developed, which is based on a sensitivity analysis of the controller parameters. This
strategy reduces directly the complexity of the optimization problem and accelerates the optimization process. Controller designs for two AMB systems are considered in this thesis. Based on the introduced and presented hierarchical evaluation strategy, the controller design for the first AMB system is obtained without specific requirements related to initial solutions. The optimal controller design is applied to a test rig with a flexible rotor supported by AMBs. The results show that the introduced optimization procedure realizes the desired results of the controlled system’s behavior. The maximal speed of 15000 rpm is reached. The second AMB system is designed for a turbo-compressor. The introduced parameter reduction strategy is applied for the controller design of this AMB system. The
controller design is optimized in the search space around an initial solution. Optimization results show the efficiency of the introduced strategy.
ER