基于PSO⁃KPCA⁃LVQ的燃气调压器故障诊断  

Gas pressure regulator fault diagnosis of based on PSO⁃KPCA⁃LVQ

作　　者：王莹[1] 王亚慧[1] 安允 WANG Ying;WANG Yahui;AN Yun(Beijing University of Civil Engineering and Architecture,Beijing 100044,China)

机构地区：[1]北京建筑大学,北京100044

出　　处：《现代电子技术》2020年第24期67-71,共5页Modern Electronics Technique

基　　金：国家自然科学基金(51471019);北京建筑大学基础研究燃气调压器故障智能诊断技术研究(X18191);北京燃气集团横向课题智能化燃气调压设施标准化及安全防护研究(H19006)。

摘　　要：针对燃气调压器故障数据的非线性、非平稳与高维度的特点,提出基于粒子群优化(PSO)的核主元分析法(KPCA)与学习向量量化神经网络(LVQ)的故障诊断方法。先采用KPCA对故障数据进行降维和降噪,由于核函数中未知参数难以确定,采用粒子群算法(PSO)进行优化得到最优的核参数;然后采用LVQ对数据进行识别分类;最后,将PSO⁃KPCA⁃LVQ算法故障诊断正确率及运行时间分别与PSO⁃KPCA⁃SVM算法以及LVQ算法进行比较。结果表明,基于PSO⁃KPCA⁃LVQ的故障诊断模型优于其他两种算法,验证了该文算法的有效性。In allusion to the characteristics of nonlinear,non⁃stationary and high⁃dimensional fault data of gas pressure regulator,a fault diagnosis method based on particle swarm optimization(PSO)⁃kernel principal component analysis(KPCA)⁃learning vector quantization(LVQ)neural network is proposed.The KPCA is used for the dimension reduction and noise reduction of fault data.As unknown parameters in kernel function are difficult to determine,the PSO is adopted to optimize and obtain the optimal kernel parameters.The LVQ is used to identify and classify the data.The fault diagnosis accuracy and running time of PSO⁃KPCA⁃LVQ algorithm are compared with PSO⁃KPCA⁃SVM algorithm and LVQ algorithm,respectively.The results show that the fault diagnosis model based on PSO⁃KPCA⁃LVQ is better than the other two algorithms,which verifies the effectiveness of the algorithm.

关 键 词：燃气调压器 故障诊断 数据处理 核参数优化 数据分类 算法比较 

分 类 号：TN131-34[电子电信—物理电子学]