EEMD结合改进PCNN模型的气体泄漏信号降噪  

EEMD Combined with Improved PCNN Model for Noise Reduction of Gas Leakage Signal

作　　者：孙烨辰 李鹏[1,2,3] 常思婕 史峰 SUN Ye-chen;LI Peng;CHANG Si-jie;SHI Feng(Jiangsu Key Laboratory of Meteorological Detection and Information Processing,Nanjing University of Information Science and Technology,Nanjing Jiangsu 210044,China;Bingjiang College,Nanjing University of Information Science and Technology,Wuxi Jiangsu 214105,China;Jiangsu Collaborative Innovation Center on Atmospheric Environment and Equipment Technology,Nanjing University of Information Science and Technology,Nanjing Jiangsu 210044,China)

机构地区：[1]南京信息工程大学江苏省气象探测与信息处理重点实验室,江苏南京210044 [2]南京信息工程大学滨江学院,江苏无锡214105 [3]南京信息工程大学江苏省大气环境与装备技术协同创新中心,江苏南京210044

出　　处：《计算机仿真》2020年第9期409-414,455,共7页Computer Simulation

基　　金：国家自然科学基金资助项目(41075115);江苏省重点研发计划社会发展项目(BE201569);无锡市社会发展科技示范工程项目(N20191008)。

摘　　要：针对气体泄漏声波信号降噪的问题,提出了一种集合经验模态分解(Ensemble Empirical Mode Decomposition,EEMD)与改进脉冲耦合神经网络(Pulse Coupled Neural Network,PCNN)相结合的降噪方法,对采集到的气体泄漏声波信号进行降噪处理,同时与EEMD和变步长自适应滤波(Least Mean Square)降噪算法进行比较。算法首先对信号做EEMD的分解,将原信号分解为9个模态分量,然后对这些分量做相干性的计算,根据分量各自的含噪情况调整参数分别做PCNN降噪,最后将信号重构。原始信号由NI仪器的cDAQ采用声传感器测得。实验结果表明,上述方法能够有效地剔除气体泄漏信号中包含的各种噪声,降噪后信噪比为16.64,均方根误差为0.0209,为后续信号分析减少了干扰,上述方法为气体泄漏声波信号的特征提取与分析提供了新的思路。Aiming at the noise reduction problem of the sound signal of gas leakage,a set empirical mode decom⁃position(EEMD)method combined with the improved pulse-coupled neural network(PCNN)was proposed to re⁃duce the noise of the collected gas leakage signal and compare it with EEMD and variable step size adaptive filtering.First,the signals were decomposed into 9 modal components by decomposes the original signal into EEMD,then the coherence of these components was calculated,PCNN noise reduction for the noise components was performed,and finally the signals were reconstructed.The original signal was measured by the cDAQ of the NI instrument using an a⁃coustic sensor.The experimental results show that this method can effectively eliminate all kinds of noises contained in gas leakage signals,the signal-to-noise ratio is 16.64 and the root-mean-square error is 0.206 after noise reduction,which improves the signal quality and provides a new idea for feature extraction and analysis of gas leakage signals.

关 键 词：降噪 集合经验模态分解 脉冲耦合神经网络 气体泄漏 预处理 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]