| Maglev technology has promising application prospects in transportation,aircraft,and industrial production as a multidisciplinary high-tech technology.Among them,maglev train is one of the hot applications of maglev technology,and the development of maglev rail transit system is an important strategic part of China’s planning and development.The most fundamental and fundamental issue with magnetic levitation technology is levitation control,which has received a lot of attention from academics.Moreover,the nonlinearity,ease of interference,and high precision requirements of the magnetic levitation system make it challenging for conventional control algorithms to suit its consistently applied.In this paper,the magnetic levitation ball system is used as a carrier to study levitation control.This paper focuses on the magnetic levitation ball system and improves the control performance of the linear auto disturbance rejection magnetic levitation controller through various methods.The following are the primary research topics covered in this essay:(1)Firstly,the basic composition and working principle of the magnetic levitation ball system are explained.Modeling the single-degree-of-freedom magnetic levitation ball system through a mechanism modeling approach to analyze the essential characteristics of the system.Based on the physical properties of the magnetically levitated sphere,the system is linearized near its operating point using Taylor’s formula.Next,the mathematical model of the system is analyzed in terms of closed-loop stability,complete controllability and exact observability.The response of the system is evaluated in various ways: on the one hand,using the dynamic response metrics of the system;on the other hand,the system response is quantified using various error integration functions.(2)Near the ideal operating point,the mathematical model of the magnetically levitated sphere system can be linearized to a second-order unstable system.In this case,the magnetic levitation sphere system as a "known" mathematical model can provide a reference for parameter tuning of the LADRC controller.In this paper,we introduce the adaptive-step cuckoo search algorithm to solve the parameter optimization problem in the LADRC controller.The LADRC controller’s parameters can be changed using the algorithm efficiently.The dynamic performance,anti-interference capability and self-adaptive capability of RBF-LADRC are better than PID and L-SMC through simulation and comparison analysis.(3)RBF neural network is introduced to map the nonlinear magnetic levitation ball system in high latitude,and the linear expression and Jacobian state equation of the system are approximated in high latitude space.To enhance the algorithm’s control performance,the LADRC controller parameters can be adaptively adjusted using the gradient descent technique.The dynamic performance,anti-interference capability and self-adaptive capability of RBFLADRC are found to be better than PID and L-SMC through simulation and comparison analysis.(4)Finally,the dynamic response performance and the following performance of both ASCS-LADRC and RBF-LADRC algorithms are verified by the Googoltech magnetic levitation system GML2001 and design experiments respectively.The experimental results show that: When the magnetic levitation ball is stably levitated,the levitation gap error of RBFLADRC is within ±0.6mm,and the levitation gap error of ASCS-LADRC is within ±1.0mm.The following performance of the algorithm is then verified,and the quantitative analysis of the algorithm response using the error integral metric leads to the conclusion that the control performance of RBF-LADRC is superior. |
