Robust control of active magnetic bearing (AMB) systems has attracted significant attention due to its ability to effectively address modeling uncertainties and enhance system robustness. However, the design of robust controllers is often time-consuming, non-intuitive, and typically dependent on trial-and-error tuning. To overcome this limitation, this paper proposes a new optimization-based framework for robust controller design. By introducing disk margin (DM) as a robustness objective together with dynamic performance objectives, the weight-function design process is transformed into an intuitive procedure for balancing the trade-off between robustness and performance. The nominal model is obtained via frequency-domain system identification method. The structures of the three weighting functions in the robust control framework are determined. Seven weighting-function parameters are then selected as optimization variables, and the necessary constraints are imposed according to the requirements of robust control and the AMB system. The robust controller is synthesized by directly selecting the weighting-function parameters from the Pareto-optimal solution set and is further validated experimentally. The results show that the designed robust controller stabilizes the system well and reduces the vibration amplitude at the first flexible mode by approximately 70% compared with the benchmark controller, thus demonstrating the effectiveness of the proposed methodology.
Zhang et al. (Sun,) studied this question.