流域梯级水电站集控中心在线自诊断系统  

Online Self-diagnosis System for Centralized Control Center of Cascade Hydropower Station in Watershed

作　　者：陈辉 饶毅 李金阳 Chen Hui;Rao Yi;Li Jinyang(Centralized Control Center of Guizhou Qianyuan Electric Power Co.,Ltd.,Guiyang 550002,China;R&d Department,Nanjing Hehai Nanzi Hydropower Automation Co.,Ltd.,Nanjing 210031,China)

机构地区：[1]贵州黔源电力股份有限公司集控中心,贵阳550002 [2]南京河海南自水电自动化有限公司研发部,南京210031

出　　处：《机电工程技术》2023年第7期182-185,共4页Mechanical & Electrical Engineering Technology

摘　　要：介绍了流域梯级水电站集控中心在线自诊断系统的结构、功能和数据治理方法。系统以生产实际需求出发,实现了状态监测、能效分析、故障诊断、运行优化、检修决策等关键功能模块,构建人机友好界面,易于使用。系统采用B/S模式架构,构建KDP平台实现与梯级电站在线监测、监控系统的数据同步及统一数据访问服务。系统对数据统一编码,建立基于冗余校验、状态关联、延时聚类的数据清洗策略,构建基于工况和事件牵引的多层次、多来源数据集成,最终实现水电多系统全景状态数据集成与治理方法。通过大数据技术、神经网络算法深入分析不同工况下水电机组部件运行数据,构建数据模型并提取特征数据作为故障判断依据,简洁易用、分析迅速,便于及时发出风险报警,实现多工况、多参数、多尺度故障在线自诊断和状态预警。该系统能有效反映机组劣化趋势,为水电企业状态检修决策、生产运行管理提供数据支持。The structure,function and data management method of online self-diagnosis system for centralized control center of cascade hydropower station in river basin are introduced.Based on the actual production demand,the system realizes the business integration,realizes the condition monitoring,energy efficiency analysis,fault diagnosis,operation optimization,maintenance decision and other key functional modules,and constructs a human-machine friendly interface,which is easy to use.The system adopts B/S mode architecture,constructs the KDP platform to realize the data synchronization and unified data access service with the cascade power station online monitoring and monitoring system.The system codes the data in a unified way,establishes the data cleaning strategy based on redundancy check,state association and delay clustering,constructs the multi-level and multi-source data integration based on the working condition and event traction,and finally realizes the panoramic state data integration and governance method of hydropower multi-system.By big data technology,neural network algorithm under different working conditions,in-depth analysis of hydropower unit operation data,building data model and extract the feature data as fault judgment,concise and easy to use,rapid analysis,to facilitate a risk alarm,realize the working condition of multi-parameter multi-scale online fault diagnosis and state warning.The system can effectively reflect the deterioration trend of the unit,and provide data support for the state maintenance decision and production operation management of hydropower enterprises.

关 键 词：流域梯级水电站 数据集成 故障诊断 状态监测 

分 类 号：TV736[水利工程—水利水电工程]