Real-time control of a magnetic levitation device based on instantaneous modeling of magnetic field

Abstract

This paper presents an approximation method to simplify the magnetic field calculation for real-time control in a magnetic levitation device. Conventional field modeling methods generally involve the numerical solution of highly nonlinear differential equations of the field. The solution of these equations will take enormous computational time and disqualifies these methods for real-time control applications. In this paper, a fast and reliable mathematical model, called pivot point, is developed for the prediction of the magnetic field distribution and verified experimentally. This model is developed based on the geometry of a magnetic stator and its magnetization pattern. The model is then used in a magnetic levitation experiment to prove its real-time capabilities in position control operations. Experiments showed that application of the pivot point method improved the positioning accuracy of the levitation system in step response and trajectory tracking and reduced the settling time.