Research On Dynamic Characteristics And Active Disturbance Rejection Control Method Of Permanent Magnetic Levitation System

Magnetic levitation technology enables objects to be levitation contactless,and the key is magnetic force offsets gravity.This contactless method avoids the adverse effects caused by friction and noise.Electromagnetic levitation technology is applied in aerospace,transportation and other fields,but continuous energization of the coil results in high heat and energy losses.In the PML（Permanent Magnetic Levitation）technology,the permanent magnet as the magnetic source of the system avoids the generation of high heat,but the control is difficult,so it is very important to study the control of the PML system.This thesis mainly studies the control of PML system.Firstly,according to the different of levitation force adjustment methods,the PML system with variable magnetic circuit and variable air gap PML system were proposed.The principle of variable magnetic circuit and variable air gap were clarifies.The dynamic model of PML system was established.The open-loop stability was judged by the method of pole distribution and step response.Aiming at the problem that the PML system has long adjustment time and easy to generate oscillation,the levitation mechanism and control characteristics of the PML system were studied.The double-loop control suitable for the PML system was proposed,so that the PML system can realize levitation more quickly and stably.Then,LADRC（Linear Active Disturbance Rejection Controller）was designed to solve the problem that the disturbance of PML system is easy to diverge.Because the traditional LADRC parameters are too large to reduce the dynamic performance of the PML system,the known information of the system was substituted into the controller for auxiliary design,thereby reducing the influence of excessive parameters.When the PML system has less known information,the dynamic characteristics of the PML system are improved smaller.Therefore,the tracking error was reconfigured based on the deviation control principle,and the improved LADRC with the model-assisted LADRC was proposed.The LESO（Linear Extended State Observer）in the improved LADRC was proved to have better convergence and better performance than traditional LESO.By setting the improved LADRC adjustment parametersω₀,ω_c and ₀b in the frequency domain and time domain respectively,the parameter adjustment law suitable for the PML system was summarized.Finally,the simulation and experiment of the PML system were carried out.In order to verify that the improved LADRC has a strong control ability for the PML system,the PID control,traditional LADRC,LADRC based on model-assisted and improved LADRC were compared.Through simulation analysis,the improved LADRC can improve the dynamic performance of the PML system.The step experimental results of the PML system with variable magnetic circuit show that the adjustment time of the improved LADRC is about0.17s.Compared with PID control,the adjusting time is shortened by 0.72s and the overshoot is reduced by 183%.The experimental results of 0.5N disturbance show that the adjustment time of the improved LADRC is about 0.72s and the displacement offset is about 0.04mm.Compared with PID control,the adjusting time is shortened by 0.72s,and the displacement is reduced by 0.225mm.The experimental results of 2Hz sinusoidal tracking show that compared with PID control,the amplitude ratio is reduced by 0.13 and the phase difference is reduced by 3.94°.Step experimental results of variable air gap PML system show that the adjustment time of the improved LADRC is about 0.38s.Compared with PID control,the adjusting time is shortened by 0.74s.The experimental results of 2Hz sinusoidal tracking show that compared with PID control,the amplitude ratio is reduced by1.21 and the phase difference is reduced by 5.62°.Based on the analysis,the improved LADRC strategy has better stability and dynamic performance for the PML system,and better solves the problem that the PML system is easy to diverge when disturbed.