小波熵在燃烧扰动相关信号提取中的应用  

Application of Wavelet Entropy on Extracting Related Signals of Combustion Disturbance

作　　者：吴颀[1] 常太华[1] 郝祖龙[1] 

机构地区：[1]华北电力大学控制与计算机工程学院,北京102206

出　　处：《华东电力》2010年第9期1436-1440,共5页East China Electric Power

基　　金：国家自然科学基金项目(50776030)~~

摘　　要：传统的机理分析无法从电厂海量数据中快速寻找到与特定燃烧扰动相关的信号,依据扰动相关信号在同一频段具有较强的波动相似性,提出了一种基于小波熵的燃烧扰动相关信号提取方法。利用小波变换的带通滤波特性得到信号的不同频段成分,再计算小波能量熵来分析信号间的相关性。以频变信号为例,分析了滑动窗口宽度和小波基的选取对小波熵的检测和识别性能的影响。对某电厂数据进行的实例分析表明,该方法不仅能有效挖掘燃烧扰动相关信号,并能区分这些信号的相关性强弱。The traditional mechanism analysis can not quickly find the specific signals related to the combustion disturbance form the massive data of power plants.According to the strong volatility similarity of the disturbance-related signals in the same frequency band,a kind of wavelet entropy based method was proposed to extract the related signals of combustion disturbance.Components of different frequency bands were firstly filtered through wavelet transforming,and then the wavelet entropy was calculated to analyze the correlation of the signals.Taking the frequency-dependent signal for example,the influence of the wavelet basis and sliding window width on the detection and identification performance of the wavelet entropy was analyzed.Example analysis of the power plant data showed that this method could not only effectively excavate correlated signals of combustion disturbance,but also distinguish the relative strength of the signals.

关 键 词：电站锅炉 燃烧扰动 相关信号 多尺度特性 小波熵 

分 类 号：TK32[动力工程及工程热物理—热能工程]