摘要
为提高核辐射探测器模拟电路的故障诊断精度,基于果蝇优化算法(FOA)和支持向量机(SVM),探讨了反向认知果蝇优化算法(RCFOA)优化SVM的核辐射探测器模拟电路故障诊断新方法。基于小波包能量提取出模拟电路故障特征集,将反向学习策略引入FOA中,提出RCFOA方法并用于SVM参数优化,提高参数设置的合理性,以优化后的SVM作为模式识别方法对特征集进行分类,得到诊断结果。测试结果表明:RCFOA-SVM法省时、诊断精度更高、更加实用有效。
In order to improve the fault diagnosis accuracy of analog circuit of nuclear radiation detector,based on fruit fly optimization algorithm(FOA)andsupport vector machine(SVM),a new fault diagnosis method of analog circuit of nuclear radiation detector optimized by inverse cognitive fruit fly optimization algorithm(RCFOA)optimize SVM was proposed.Based on the wavelet packet energy,the fault feature set of analog circuit was extracted.Then the reverse learning strategy was introduced into FOA and RCFOA was proposed and applied to the parameter optimization of SVM to improve the rationality of parameter setting.The optimized SVM was used as the pattern recognition method to classify the feature set and obtain the diagnosis results.The test results show that the RCFOA-SVM method saves time,has higher diagnostic accuracy and is more practical and effective.
作者
谭鹤毅
张伟
闵丙源
TAN He-yi;ZHANG Wei;MIN Bing-yuan(Department of Electronic Information Engineering,Nanchong Vocational and Technical College,Nanchong Sichuan 637131,China;School of Automation Engineering,University of Electronic Science and Technology,Chengdu 611731,China;College of Engineering,Mokwon University,Daejeon 35349,Korea)
出处
《核电子学与探测技术》
CAS
北大核心
2022年第4期646-651,共6页
Nuclear Electronics & Detection Technology
基金
国家自然科学基金(61201131)
国家高技术研究发展计划(2011AA090101)
“十三五”国家科技重大专项(2017ZX05019-002)资助。
关键词
反向认知果蝇优化算法
支持向量机
故障诊断
模拟电路
核辐射探测器
Fruit Fly Optimization Algorithm with Reverse Cognition
Support Vector Machine
Fault Diagnosis
Analog Circuit
Nuclear Radiation Detector
