Robust output tracking of transverse flux machines using RBF neural network

This paper presents an application of radial basis function (RBF) in the identification and control design of transverse flux machines as nonlinear systems with unknown nonlinearity part. The technique of feedback linearization and H/sub /spl infin// control are used to design an adaptive control law for compensating the unknown nonlinearity part, such that the effect of the cogging torque as a disturbance is decreased into the angle and angular velocity tracking.