水轮发电机轴电流超标故障自动诊断技术研究  

Research on Automatic Diagnosis Technique of Hydro Generator Axial Current Exceeding Faults

作　　者：刘立春 LIU Lichun(Guoneng Daduhe Zhentouba Hydropower Construction Co.,Ltd.,Leshan 614000,China)

机构地区：[1]国能大渡河枕头坝发电有限公司,四川乐山614000

出　　处：《自动化仪表》2025年第2期97-101,共5页Process Automation Instrumentation

摘　　要：水轮发电机故障类型多样且复杂。为了在规定的运行范围内分析振动信号变化、准确诊断水轮发电机轴电流超标故障、提升设备运行安全性,提出水轮发电机轴电流超标故障自动诊断技术。通过独立分量分析-互补集合经验模态分解方法,重构振动信号。应用经验小波变换方法,求解轴电流超标故障特征矢量。选择经验小波变换分量提取轴电流超标故障引起的振动信号特征。使用遗传-粒子群算法优化径向基函数(RBF)神经网络模型,计算粒子的最优适应度和最佳位置,以自动诊断轴电流超标故障。试验结果表明,所提技术的水轮发电机轴电流超标故障自动诊断准确率高。所提技术能够提升设备运行安全性,具有较高实际应用价值。Hydro generator fault types are diverse and complex.To analyze the changes of vibration signals,accurately diagnose the hydro generator axial current exceeding faults,and improve the safety of equipment operation within the specified operation range,the automatic diagnosis technique of hydro generator axial current exceeding faults is proposed.The vibration signal is reconstructed by the independent component analysis-complementary ensemble empirical modal decomposition method.The empirical wavelet transform method is applied to solve the feature vector of axial current exceeding fault.The empirical wavelet transform components are selected to extract the vibration signal features caused by the axial current exceeding fault.The radial basis function(RBF)neural network model is optimized using a genetic-particle swarm algorithm to calculate the optimal fitness and the best position of the particles to automatically diagnose the axial current exceeding fault.The experimental results show that the proposed technique has high accuracy in the automatic diagnosis of axial current exceeding faults of the hydro generator.The proposed technique can improve the safety of equipment operation and has strong practical application value.

关 键 词：水轮发电机 轴电流超标 经验小波变换 遗传-粒子群算法 最优适应度 径向基函数神经网络 

分 类 号：TH-39[机械工程]