基于时序深度融合网络的汽轮发电机组振动故障诊断  

Vibration Fault Diagnosis System Based on Timing Deep Integration Network

作　　者：何小锋[1] HE Xiaofeng(Jiangsu Frontier Electric Technology Co.,Ltd.,Nanjing 211102,China)

机构地区：[1]江苏方天电力技术有限公司,江苏南京211102

出　　处：《山东电力技术》2024年第10期67-73,共7页Shandong Electric Power

基　　金：国家重点研发计划项目(2022YFB4100403)。

摘　　要：将深度学习智能算法应用到汽轮发电机组振动故障诊断领域,推进汽轮发电机组振动故障智能诊断的进步,采用深度学习结合专家经验的方法,根据振动专家现场振动故障诊断的经验,将振动故障的时序特征及运行参数对故障的影响融入传统的深度学习算法中,提出了基于时序深度融合网络的振动故障诊断算法。通过将时序特征及运行参数等对振动故障诊断有重要影响的静态量数据适当处理变换,使其能够应用到深度学习算法模型中,研究了该诊断系统的相关关键技术,包括特征提取、网络搭建、异构信息融合等。实验数据验证结果表明,融合了时序特征及运行参数信息的深度学习诊断算法在提高故障诊断准确率的同时,大大提升了网络性能,提高了网络鲁棒性和迁移能力。In order to apply the latest deep learning intelligent algorithm in the field of vibration fault diagnosis of steam turbine generator sets and promote the progress of intelligent diagnosis of vibration faults of steam turbine generator sets,the method of combining deep learning with expert experience is adopted,and according to the experience of vibration experts in on-site vibration fault diagnosis,the timing characteristics of vibration faults and the impact of operating parameters on faults are integrated into the traditional deep learning algorithms,and a vibration fault diagnosis algorithm based on the time series deep integration network is proposed.By appropriately processing and transforming static data such as timing characteristics and operating parameters that have a significant impact on vibration fault diagnosis,they can be applied to deep learning algorithm models.The relevant key technologies of the diagnostic system were studied,including feature extraction,network construction,heterogeneous information fusion,etc.Experimental data verification results show that the algorithm greatly improves network performance,network robustness and migration ability while improving the accuracy of fault diagnosis.

关 键 词：深度学习 时序 运行参数 振动 汽轮机 

分 类 号：TM62[电气工程—电力系统及自动化] TP18[自动化与计算机技术—控制理论与控制工程]