【作者】 赵俊龙； 郭正刚； 张志新； 李宏坤； 王奉涛；

【Author】 Zhao Junlong Guo Zhenggang Zhang Zhixin Li Hongkun Wang Fengtao (Key Laboratory for Precision and Non-traditional Machining Technology of Ministry of Education, Dalian University of Technology, Dalian, 116023, China)

【机构】 大连理工大学精密与特种加工教育部重点实验室；

【摘要】 往复式压缩机组合阀信号的时域包络曲线包含了丰富的故障信息,可以用来表征组合阀的工作状态。提出了一种基于小波包降噪、包络分析以及支持向量机的组合阀工作状态识别方法。首先应用小波包降噪方法对信号进行降噪处理,其次对降噪后信号进行包络分析,然后从包络曲线中提取峰值特征指标构造支持向量机的特征向量空间,最后应用多分类支持向量机实现组合阀工作状态的智能识别,工程应用表明该方法可以有效的识别出往复式压缩机组合阀的各类工作状态。更多还原

【Abstract】 The envelope of reciprocating compressor combined valve signal,which carries plenty of useful fault information,is proved to be very useful in reciprocating compressor working condition identification.A new method for combined valve working condition identification based on wavelet-packet denoise,envelope analysis and support vector machine is presented.First,wavelet-packet is used to reduce noise from the signal.Second,the envelope of the denoised signal is obtained.Finally,envelope peak amplitude indexes are extracted to construct eigenvectors,which are used to identify different working conditions of combined valve by using multi-class support vector machine.The engineering application verifies its effectiveness in fault classification of combined valve signals.更多还原