Auto-Disturbance-Rejection Control Of Maglev System

The levitation system of maglev train has strong nonlinearity and model uncertainty. Traditional control arithmetics mostly rely on precise object model, which show limited capacity on restricting some certain external disturbance and variation of system parameters. Especially when a certain levitation point is breakdown, the point nearby will endure enormous sudden-varied load due to mechanical coupling, in which the controller is required the strong ability of rejecting the disturbance. Therefore, it is of great significance to figure out a levitation control method with the ability of strong disturbance rejection other than depending on the system model.ADRC is a control method based on disturbance rather than relying on the object model, which can ensure control precision on the condition of strong nonlinearity and uncertain strong disturbance. At present ADRC mainly consists of TD, ESO and NLSEF. Basing on ADRC and setting the single point levitation system as the object, two control methods have been designed, namely, nonlinear PID and double-loop ADRC. Subsequently, several regulations for parameters'tuning are presented. The simulation results demonstrate that both methods have advantages of quick response and no overshoot, of which the latter has stronger robustness and adaptability as to massive disturbance of external system and internal parameters'variation. In addition, the ability of disturbance rejection is enhanced when the model supplement is applied in the double-loop ADRC.When comes to analyze the closed-loop stability, which isn't mature in theory, the nonlinear layout of ADRC, along with the estimate and supplement for the extended state of ESO, brings in abundant difficulties. Thus, based on simplified layout of ADRC, the analysis on the closed-loop stability of the first-order ADRC for current-loop has been conducted. Nonlinear layout is widely used in ADRC due to its better capacity of disturbance rejection and astringency than linear structure. Taken this into consideration and based on the TD and ESO with a new kind of TD introduced, the simulation research of filtering and differential is conducted, providing reference for obtaining damp from clearance differential.At last, the hardware platform of levitation control is constructed and relevant program is presented. When the experiment on the techniques of ADRC has been conducted, the results illustrate that it is efficient to arrange the transitional process, filter the signals of clearance and acceleration, and extract the clearance differential by TD. Meanwhile, ADRC can realize the stable levitation of single point.