采用人工神经网络法对电器触头的逆问题求解  

Investigation on inverse problem of circuit breaker using artificial neural networks

作　　者：刘冬[1] 曹云东[1] 杜威[1] 孙静[2] 

机构地区：[1]沈阳工业大学电气工程学院,辽宁沈阳110023 [2]沈阳产品质量检验研究所,辽宁沈阳110021

出　　处：《沈阳工业大学学报》2004年第2期160-163,共4页Journal of Shenyang University of Technology

摘　　要：提出了将人工神经网络与电磁场数值分析相结合进行电器电场优化设计的一种新方法.利用人工神经网络的合理结构得到了非线性逼近解,并结合电磁场数值分析计算的准确性,实现了电器结构优化设计逆问题求解的快速性和准确性.实例计算结果证明了人工神经网络(ANN)在优化设计中的可行性和有效性.A combination of artificial neural networks (ANN) and the electric field analysis is applied to optimize the electric field. Taken the high nonlinear approaching ability of ANN and the numerical analysis of electromagnetic field into consideration, the inverse problem solution with quick and accurate characteristic has been performed. From the calculation results, it can be seen that the proposed method is feasible and valid.

关 键 词：逆问题求解 人工神经网络 仿真 

分 类 号：TM561.3[电气工程—电器]