This paper presents the improved characteristic analysis technique for the
5-phase hybrid stepping motor used for servo systems. The basic approach is
based on the equivalent magnetic circuit method not only to deal with a 3-
D nature of the motor but also to save the analysis time of the motor. And,
the Finite Element Method (FEM) is used to generate the circuit parameters
considering the saturation effects in teeth and poles. In addition, the ne
ural network is introduced to model the variation of the parameters and app
lied to predict the unknown parameter values. This paper demonstrates the e
ffectiveness of the improved method by computer simulation and experimental
results.