Data-driven control loop performance evaluation of electromagnetic levitation systems

Electromagnetic levitation system is the core component of maglev trains, and the performance of its control loop directly affects the stability, safety and riding comfort. After a long time of operation, the control loop performance of the electromagnetic levitation system will be reduced from its initial condition, due to changes in external environment, actuator failure, etc. Therefore, monitoring and evaluating the control loop performance is of great significance to ensure a smooth operation of the maglev train. In this paper, the concept and framework of control loop performance evaluation are introduced to the electromagnetic levitation system for the first time; and then the fundamental for control loop performance evaluation based on data-driven methods is presented; with the operational data of commercial maglev train, performances of different evaluation methods are tested and compared. In addition, by taking advantages of the geometrical significance of covariance matrix, evaluation results of the control loop performance can be visualized in a compact and intuitive fashion.