摘要
提出了一种基于本征模态函数能量距法对乏燃料对剪切机工作噪声进行特征提取,并融合了BP神经网络和支持向量机构建了一种混合模型用于乏燃料剪切机刀具磨损状态的监测。对乏燃料剪切机刀具的正常、轻度磨损、重度磨损和损坏四种状态下的工作噪声信号分析的结果表明,该方法可以准确、有效地识别这些状态。
In this paper,a feature extraction method based on intrinsic mode function（IMF） energy moment is presented for extracting the features of noise produced by the working shearing machines.And,by combining the BP neural network（BPNN） and support vector machine（SVM）,a hybrid BPNN-SVM model is proposed for tool wear condition monitoring of spent fuel shearing machines.Empirical study on the working noise samples of the spent fuel shearing machine under four tool wear states（normal,mild wear,severe wear and damage） is carried out.The results show that the method can identify these states accurately and effectively.
作者
陈甲华
邹树梁
CHEN Jia-hua;ZOU Shu-liang(School of Nuclear Science and Technology,University of South China,Hengyang Hunan 421001,China;School of Management,University of South China,Hengyang of Hunan Prov.421001,China;Hunan Provincial Key Laboratory of Emergency Safety Technology and Equipment for Nuclear Facilities,University of South China,Hengyang of Hunan 421001,China)
出处
《核电子学与探测技术》
CAS
北大核心
2018年第2期298-303,共6页
Nuclear Electronics & Detection Technology
基金
湖南省军民融合产业发展专项(2013JMH01)
湖南省科技厅重点研发项目(2015GK3030)资助
关键词
乏燃料剪切机
刀具磨损
状态监测
IMF能量距
BP神经网络
支持向量机
spent fuel shearing machine
tool wear
condition monitoring
IMF energy moment
BP neuralnetwork
support vector machine
