Modeling and optimization of the claw-pole alternator based on support vector machines and chaos

A novel way about nonlinear regression modeling of the claw-pole alternator oriented to the computer-aided design (CAD) with the support vector machines (SVM) is presented in this paper. The training samples are rooted in the severe finite element model (FEM) simulation. Parameter optimization of the claw-pole alternator is also presented, which is based on chaos and uses the SVM regression model as the objective function. This integration preserves the benefits of the SVM regression model and the chaos, where the SVM regression model is high efficiency and the chaos might be very useful for finding the optimal structure parameter.