基于Volterra核的MIMO非线性电路建模及智能特征提取  

Model building and intelligent feature extraction for MIMO non-linear circuit based on volterra kernel

作　　者：陈叶 廖耀华 王恩 朱梦梦 李博 陈寅生 林海军[2] Chen Ye;Liao Yaohua;Wang En;Zhu Mengmeng;Li Bo;Chen Yinsheng;Lin Haijun(Electric Power Research Institute of Yunnan Power Grid Co.,Ltd.,Kunming 650217,China;Harbin University of Science and Technology,Harbin 150080,China)

机构地区：[1]云南电网有限责任公司电力科学研究院,昆明650217 [2]哈尔滨理工大学,哈尔滨150080

出　　处：《电测与仪表》2021年第10期170-176,共7页Electrical Measurement & Instrumentation

基　　金：国家自然科学基金资助项目(61803128);中国南方电网有限责任公司科技项目(YNKJXM20170549)。

摘　　要：为了解决模拟乘法器等多输入测量电路的智能故障诊断准确率低的问题,文中研究了多输入多输出(MIMO)电路的基于Volterra级数的建模方法,为电路的故障诊断提供模型,提出了整体退火遗传特征提取方法,利用整体退火遗传算法的全局寻优能力优化故障诊断特征参数的提取,以选出各种故障状态之间特征差异最大的特征,以提高故障诊断的准确率,以模拟乘法器电路为例进行了建模及故障特征智能优化提取实验。实验表明,文中方法可以有效建模并提高智能故障诊断的准确率。In order to solve the problem of low accuracy in intelligent fault diagnosis for multi-input measuring circuit such as analog multiplier,this project studies the model building method for multi-input multi-output(MIMO)circuit based on series of Volterra,which provides the model for circuit fault diagnosis.Then,the method of feature extraction for whole annealing genetic features is proposed by utilizing the global optimization property of whole annealing genetic algorithm(WAGA)to optimize the parameter extraction for fault diagnosis feature.Then,the feature with the largest feature difference between various fault status is selected to improve the accuracy of fault diagnosis.Finally,we conduct an experiment of intelligent optimization and extraction of model-building features and fault features through using analog multiplier as an example.The experimental result has proved that,the method proposed in this paper is effective in model-building and improving the accuracy of intelligent fault diagnosis.

关 键 词：多输入多输出电路 VOLTERRA级数 整体退火遗传算法 智能特征提取 故障诊断 

分 类 号：TM93[电气工程—电力电子与电力传动]