摘要
在尽量简化取样的条件下实现自动故障诊断,可以通过对电路输出进行傅立叶分解得到量化的信息,基于粗糙集理论构造的分类器适用于这种场合,因其具有在输入不精确、不完整信息情况下的分类能力;在傅立叶分解、数据离散化、采用RSES进行训练的基本流程中,取样点的选取,离散算法等细节对故障的识别准确率有较明显的影响;实验结果证明了粗糙集理论结合傅立叶分解诊断电子电路故障的可靠性和准确性。
Rough set theory is introduced into Auto-fault-detection of analog circuits. The output signal analysis of Fourier-transform can expose several features of fault type, but which sample point should be selected and how to discretize these analog signals？ Such problems is worthy enough to research with the purpose of improving the classification accuracy of Rough Set method.
出处
《计算机测量与控制》
CSCD
2007年第6期712-713,725,共3页
Computer Measurement &Control
关键词
粗糙集
故障诊断
模拟电路
傅立叶分解
rough set
fault detection
analog circuit
Fourier-transform
