基于Volterra级数的水轮机调速器液压系统故障诊断  

Application of Fault Diagnosis Approach Based on Volterra Series for Hydraulic System of Turbine Governor

作　　者：付一丁[1] 李穆[2] 李丰攀[2] 顾然[2] 

机构地区：[1]中国长江三峡集团公司,北京100091 [2]国网电力科学研究院武汉南瑞有限责任公司,湖北武汉430074

出　　处：《水力发电》2016年第2期64-69,105,共7页Water Power

摘　　要：针对现有水轮机调速器液压系统故障诊断存在诸如诊断精度不高、泛化能力不强、在线检测能力不足等问题,根据Volterra级数的非线性模型高辨识能力以及可实时在线计算等优点,本文提出了一种基于Volterra级数的水轮机调速器液压系统故障诊断的新方法。该方法利用递推最小二乘法辨识出系统Volterra核,经多维广义离散Fourier变换得到系统广义频域响应函数(GFRF)模型,最后利用主元分析法(PCA)进行特征提取及分类,从而实现系统故障的高效诊断。仿真结果表明,该方法具有较高辨识精度,能准确分辨出系统的不同状态。At present, there are some problems in fault diagnosis of hydraulic system of hydraulic turbine governor, such as poor model identification ability, low generalization ability and limited on-line detection ability. By using the high identification ability in nonlinear model and the on-line calculation ability of Voherra series, a new fault diagnosis method is put forward. Firstly, the method uses the recursive least-squares to identify the Voherra kernel of system. Secondly, the Discrete-time Fourier Transform （DTFT） is applied to gain the Generalized Frequency Response Functions （GFRF） model of system. Finally, the characteristics which being extracted from the GFRF are pushed into Principal Component Analysis （PCA） module to classify the different status of system in a coordinate grid map. The simulation results show that this method has a clear process, a powerful adaptable ability and strong identification ability, which can accurately distinguish different system status.

关 键 词：VOLTERRA 广义频域响应函数(GFRF) 主元分析法(PCA) 故障诊断 调速器 水轮机 

分 类 号：TV136[水利工程—水力学及河流动力学]