基于大数据的水电站运维远程智能监控系统构建与应用  

Construction and Application of Remote Intelligent Monitoring System for Hydropower Station Operation and Maintenance Based on Big Data

作　　者：吴新宇 WU Xinyu(Sichuan Liangshan Shuiluohe Electric Power Development Co.,Ltd.,Chengdu 610095,China)

机构地区：[1]四川凉山水洛河电力开发有限公司,四川成都610095

出　　处：《中国水能及电气化》2024年第6期8-13,28,共7页China Water Power & Electrification

摘　　要：为解决偏远山区水电站运行维护难的问题,基于大数据、云服务、传感器和人工智能技术,设计开发了一套水电站运维远程智能监控系统,系统通过传感器对现场设备进行监测,然后通过局部均值分析法(LMD)对现场采集到的复杂振动信号进行分解处理,最后利用Elman神经网络的故障诊断专家系统进行故障诊断,可实现远程监测、控制和故障诊断等功能;系统在设计的100人并发用户下,CPU占有率仅为11.7%,每月消耗网络流量395.8GB<1000GB(设计网络流量),平均每一次响应时间仅为3.3ms;系统大大降低了运维人员的工作强度,提高了故障诊断效率,降低了运维成本,在实际工程中取得了较好的应用效果。In order to address the challenges of operation and maintenance in remote mountainous hydropower stations,a remote intelligent monitoring system for hydropower station operation and maintenance is designed and developed based on big data,cloud services,sensors,and artificial intelligence technology.The system monitors on-site equipment through sensors and then decomposes and processes the complex vibration signals collected on-site using the Local Mean Decomposition(LMD)method.Finally,fault diagnosis is performed using the Elman neural network expert system,enabling remote monitoring,control,and fault diagnosis functions.Under the design condition of 100 concurrent users,the system has a CPU occupancy rate of only 11.7%and consumes 395.8GB of network traffic per month,well below the designed network traffic of 1000GB.The average response time is only 3.3ms.The system significantly reduces the workload of maintenance personnel,improves fault diagnosis efficiency,and reduces maintenance costs,achieving good application results in practical engineering.

关 键 词：水电站运维 远程智能监控系统 局部均值分析法 ELMAN神经网络 故障诊断 

分 类 号：TV736[水利工程—水利水电工程]