Study of Taper Roll-pole Type Magnetic Suspension Bearing Control Using Fuzzy Neural Network

摘要

The test system of TRTMSB is introduced and its mathematical model is established using lineal control theory. The nonlinear controller is designed based on fuzzy parameters and rules taken from a linear optimal controller. Variables after fuzzy are incorporated in the neural network structure. Neural network parameters are optimized by the BFGS method and the feedback gain matrix is obtained after de-fuzzification. Simulation results show that the controller can improve both steady stability and transient stability.