河北水利电力学院学报 ›› 2023, Vol. 33 ›› Issue (1): 19-23.DOI: 10.16046/j.cnki.issn2096-5680.2023.01.004

• 电气工程专题 • 上一篇    下一篇

基于概率神经网络的水轮机组水力振动故障诊断

苏立1, 毛成1, 沈春和1, 谢文经2, 戴利传2   

  1. 1.贵州电网有限责任公司电力科学研究院,贵州省贵阳市解放路32号 550002;
    2.贵州黔能企业有限责任公司,贵州省贵阳市市南路120号 550000
  • 收稿日期:2022-03-23 修回日期:2022-07-05 发布日期:2023-10-26
  • 作者简介:苏 立(1982-),男,山西太原人,高级工程师,主要研究方向为水电机组仿真及故障诊断。E-mail:docsuli198411@163.com
  • 基金资助:
    贵州省科技支撑计划项目(黔科合支撑[2020]2Y042)

Fault Diagnosis of Hydroturbine Hydraulic Vibration Based on Probabilistic Neural Network

SU Li1, MAO Cheng1, SHEN Chun-he1, XIE Wen-jing2, DAI Li-chuan2   

  1. 1. Electric Power Research Institute of Guizhou Power Grid Co., Ltd., 550002, Guiyang,Guizhou, China;
    2. GuizhouQianneng Enterprise Co., Ltd., 550000, Guiyang,Guizhou, China
  • Received:2022-03-23 Revised:2022-07-05 Published:2023-10-26

摘要: 故障诊断技术是水电站水轮机组安全稳定运行的关键技术之一。针对常规在线监测系统难以发现水轮机组振动故障问题,提出了一种基于概率神经网络的水轮机组故障诊断模型。该模型主要由故障样本数据预处理、样本数据归一化和概率神经网络等三个部分组成。诊断结果表明,所诊断样本与实际的故障类型基本一致,具有良好的诊断效果。

关键词: 水力振动, 故障分类, 概率神经网络, 水轮机组, 安全运行

Abstract: Fault diagnosis technology is one of the key technologies for the safe and stable operation of hydroelectric turbine units in hydropower plants. As the conventional online monitoring system is difficult to detect the unit vibration fault problem, a probabilistic neural network (PNN) algorithm is constructed, and the fault samples are designed for the fault parameters that have been collected. The program mainly includes three parts such as fault sample data pre-processing, sample data normalization and probabilistic neural network. The results show that the diagnostic results of the PNN model are consistent with the actual fault types, and this model has good diagnostic effects.

Key words: hydraulic vibration, failure types, PNN, water turbine, running safety

中图分类号: