Sliding Mode Control Based On Fuzzy Nearal Network On Maglev Guiding System

As maglev guiding technology has advantages such as low noise, no friction and, etc.it has been applied in linear elevator that makes the guide shoes and the guiding rail nocontacting to eliminate mechanical friction and achieve high-speed operation. Linearelevator is an important development direction in the application of the tall building in thefuture. A new guiding structure of elevator is presented, the main research is when linearelevator startup, air-gap control in x axis system of single maglev guiding system in linearelevator inhibitions the effect of the parameter uncertainty and external disturbance on theair gap, and ensures achieve fast tracking and strong robustness.First, the background and development of maglev technology and application-orientedsituation is reviewed, meanwhile, a brief overview of magnetic guiding technology andcontrol method in linear elevator. For the problem of maglev guiding system in linearelevator, the permanent magnet and inductor are added in "Eight-Maglev" topology, at thesame time, using electric (EDS) and electromagnetic (EMS) hybrid suspension system, andputting forward overall control method of linear elevator.Secondly, mathematic model of single magnetic levitation guiding system isestablished when the linear elevator is starting. Because the system mathematical model isnonlinear, so using equilibrium point linearization method to make control systemmathematical model simplified, achieving the linear model of current and voltagecontrolling, and using current control model as the mathematical model of the controller.Since single maglev guiding system has the characteristics of fast-track and immunity, theintegral sliding mode control method is used. According to the chattering problem ofsliding mode structure control, introducing adaptive control estimates gain of sliding modeswitch control quantity, real-time adjusting sliding mode control to eliminate chattering.For the jittering problem of elevator when system parameter is changing, using the fuzzyneural network instead of sliding mode equivalent control to realize accelerating response,improving the robustness, weakening the vibration of the elevator. Finally, the Matlab simulation software is used to analyze the control strategy aboveon, the simulation results make it is obvious that the fuzzy neural network (FNN) slidingmode control can realize fast tracking performance and strong robustness of single maglevguiding system when linear elevator is starting.