Control of Magnetic Suspended Flywheel using adaptive linear neuron

Focused on robustness, low power consumption and unbalance compensation demands from Magnetic Suspended Flywheel (MSF), a network controller is presented using on adaptive linear neuron, and characteristic equation is derived and discussed form the point of stability, then, a method to ensure close loop stability is established by checking the update of neural network weight. Simulations are performed based on MSF nonlinear model, the results indicate that rapid response, low power consumption and robustness is achieved by the adaptive linear neuron control, unbalance vibration is eliminated under the constraints of power consumption, besides, the stability of close loop system is guaranteed by the weight update checking method.