基于有向无环图支持向量机的水轮发电机组故障诊断模型.pdfVIP

第34 卷 第2 期 电 网 技 术 Vol. 34 No. 2 2010 年2 月 Power System Technology Feb. 2010 文章编号：1000-3673 （2010 ）02-0115-05 中图分类号：TM 612 文献标志码：A 学科代码：470·4051 基于有向无环图支持向量机的 水轮发电机组故障诊断模型 兰飞，唐玲 （广西大学电气工程学院，广西壮族自治区 南宁市 530004 ） Fault Diagnosis Model for Hydropower Generating Unit Based on Directed Acyclic Graph Support Vector Machine LAN Fei, TANG Ling (School of Electrical Engineering, Guangxi University, Nanning 530004, Guangxi Zhuang Autonomous Region, China) ABSTRACT: It is significant for secure operation of power [2-3] [4-6] [7-8] Petri 网 、小波分析 、信息融合技术 、专家 grids to improve both accuracy and timeliness of status 系统[9-10]和人工神经网络[11-14]等，其中基于神经网 monitoring and fault diagnosis system for hydropower 络的水轮机组故障诊断模型最常用。神经网络算法 generating sets. To remedy the defects of neural network such 是一种较好的智能分类算法，能正确地判断机组的 as slow convergence rate and easy to sink into locally optimal 故障类型。然而收敛速度慢、容易陷入局部最小等 solution, a fault diagnosis method for hydropower generating sets, which is based on directed acyclic graph support vector 缺点限制了其在实际工程中的应用。基于此，本文 machine, is proposed. The proposed method can effectively 研究一种基于频谱分析和支持向量机方法的水轮 solve such problems as small sample, high dimension and 发电机新模型，以求快速准确地对故障做出响应。 nonlinearity, thus the fault of hydropower generating sets can 支持向量机(support vector machine，SVM)是一 be accurately diagnosed in a shorter time period. 种以有限样本统计学习理论为基础的新通用学习 KEY WORDS: hydropower generating uni