基于EEMD⁃PCA⁃OSELM的燃气调压器故障诊断  

Gas pressure regulator fault diagnosis using EEMD⁃PCA⁃OSELM method

作　　者：李胜楠 王怀秀 王亚慧 宋洋 LI Shengnan;WANG Huaixiu;WANG Yahui;SONG Yang(School of Electrical and Information Engineering,Beijing University of Civil Engineering and Architecture,Beijing 100044,China)

机构地区：[1]北京建筑大学电气与信息工程学院,北京100044

出　　处：《现代电子技术》2022年第9期137-142,共6页Modern Electronics Technique

基　　金：国家重点研发计划支撑项目(2018YFC0807806);国家自然科学基金资助项目(51471019)。

摘　　要：燃气调压器存在故障数据样本少、发生故障不易察觉等问题,传统的离线诊断模型难以有效学习故障数据的特征信息且难以实时更新诊断系统。针对上述问题,提出一种基于集合经验模态分解(EEMD)、主元分析法(PCA)与在线贯序极限学习机(OSELM)结合的故障诊断方法。利用EEMD对获取的故障数据流进行频域分解,并通过PCA对已分解的不同频率分量进行特征提取;然后,随机选取少量经处理后的故障特征样本利用极限学习机(ELM)算法对模型进行初始化,并将剩余样本经EEMD⁃PCA处理后以数据流的方式对现有模型进行更新,通过在线增量学习方法递推计算故障诊断系统参数并给出诊断决策。利用某调压器故障信息进行仿真实验,结果表明,所提EEMD⁃PCA⁃OSELM故障诊断方法能在保证较高识别率的前提下实现快速故障诊断。There are some problems existing in gas pressure regulator,for example,samples of fault data is few,the failure is not easy to detect.However,the traditional offline diagnosis model is far away from effectively learning the feature information of fault data and updating the diagnosis system in real time.In view of the above,a fault diagnosis model is established based on ensemble empirical mode decomposition(EEMD),principal component analysis(PCA)and online sequential extreme learning machine(OSELM).The acquired fault data stream is subjected to frequency domain analysis by EEMD.Different decomposed frequency components are subjected to feature extraction by PCA.A small number of processed fault feature samples are selected at random.The extreme learning machine(ELM)algorithm is used to initialize the model.After the remaining samples are processed by EEMD⁃PCA,the existing model is updated by data stream.The parameters of fault diagnosis system are recursively calculated by online incremental learning method,and the diagnosis decision is given.The fault information of a certain regulator is used for simulation experiment.The results show that the proposed method can achieve faster fault diagnosis with higher recognition rate.

关 键 词：在线贯序极限学习机 故障诊断 燃气调压器 集合经验模态分解 主元分析法 极限学习机 增量学习 数据流 

分 类 号：TN99-34[电子电信—信号与信息处理] TU996[电子电信—信息与通信工程]