Nonlinear control of magnetic suspension system based on RBFNN

In this paper, nonlinear characteristic of a multivariable magnetic suspension system is formulated, and a modified scheme of nonlinear multivariable internal model control (IMC) is put forward, based on neural network inversion. With radial basis function neural network (RBFNN) approximating inversion of the system, the compound system is linearised and decoupled, and IMC can be applied here for this nonlinear magnetic suspension system. In the end, simulations can prove the effectiveness of the improved approach.