【摘要】 通过电机参数样本空间设计,引入支持向量机,对爪极发电机的电磁模型进行非线性回归建模分析,基于混沌理论对爪极发电机结构参数进行优化.仿真结果表明,支持向量机用于爪极发电机非参数建模准确可行,并且是高效的,非常适合于需要大规模迭代计算的参数优化.将有限元电磁仿真与支持向量机结合用于非参数建模,以及在非参数模型的基础上用混沌进行优化,这为爪极发电机以及其它的电磁工程设计提供了一种新的思路.更多还原

【Abstract】 Support vector machines(SVM) is successfully applied in nonlinear regression modeling of the claw-pole alternator by designing the sample space of the parameters in this paper.Furthermore,parameters optimization of the claw-pole alternator based on chaos is also introduced.By comparing the simulation results with finite element model(FEM),the nonparametric modeling based on SVM is proved to be feasible and highly efficient.Hence,it is quite fit for the parameters optimization that needs large-scale iterative calculations.Thus,integration of SVM and the traditional FEM in nonparametric modeling and parameters optimization based on chaos are presented and verified.These efficient methods can be used in the optimum design of the claw-pole alternator and other electromagnetic engineering.更多还原