摘要
提出了一种开关磁阻电机(SRM)数学建模的新方法.在已知SRM 5个特殊位置电感数据的基础上,将傅里叶级数分解和曲线拟合方法运用于磁链模型的建立中,根据虚位移原理计算获得SRM的矩角特性,并将自适应神经网络模糊推理系统(ANFIS)用于矩角模型的建立.基于非线性磁链模型与ANFIS矩角模型,对一台6/4结构的SRM进行了仿真与实验.仿真结果与实验结果基本一致,最大误差不超过5%,从而验证了这种建模方法的正确性.同时,该建模方法还可以进一步应用于SRM的磁链控制和转矩控制中,为工程设计和调试提供依据.
A novel mathematic model for switched reluctance motor（SRM）was proposed. The Fourier series expression and curve fitting method were used to model the flux linkage according to the inductance data of five positions, then the torque characteristics were sought out via the principle of virtual displacement, and adaptive network-based fuzzy inference system （ANFIS） was utilized to train the torque-angle model. The simulation and experiment were carried out comparatively on SRM with six stator poles and four rotor poles based on the nonlinear flux linkage model and torque-angle model of ANFIS to demonstrate the both coincidence with a maximal error lower than 5%. And the method is expected to apply to SRM flux linkage control and torque control to provide the basis for engineering design and debugging.
出处
《西安交通大学学报》
EI
CAS
CSCD
北大核心
2007年第2期214-218,共5页
Journal of Xi'an Jiaotong University
基金
陕西省科学技术研究发展攻关计划资助项目(2005K07-G9)
关键词
开关磁阻电机
磁链模型
矩角模型
自适应神经网络模糊推理系统
switched reluctance motor
flux linkage model
torque-angle model
adaptive network- based fuzzy inference system
