摘要
针对目前水电机组状态监测系统中广泛采用的单一阈值预警模式的局限性,探寻更完善的预警策略,以对机组特性的波动及变化进行监测和预警。通过对实际的监测数据进行概率统计分析,利用数学工具对其概率分布进行了严格的分布检验,结果表明,监测参数的样本均值序列服从正态分布。进而利用正态分布的理论知识,提出了基于3σ准则的水电机组监测参数波动预警策略。相对于常见的单一阈值的越限预警模式,波动预警模式可快速有效地监测机组偶发的小幅变化,做出及时预警,有助于运行人员及时发现机组正在发生的渐变的运行特性变化,提高监测和预警的有效性,保障机组运行安全。
Due to the limitation of out-of-limit strategy,fluctuation and variation of unit characteristics is difficult to be detected by the condition monitoring system of hydropower unit.Probability statistics of monitoring parameters of hydropower unit is performed in this paper.Strict distribution test is carried out on the probability distribution by specific mathematical tool.Test results show that average sequence of sampling data obeys the Gaussian distribution.Strategy of fluctuating alarm for hydropower parameters is proposed by 3σcriterion based on Gaussian distribution theory.Compared to the general out-of-limit strategy,this strategy can monitoring the minor fluctuation of parameters sensitively and detect the gradually varied characteristics of the unit.This fluctuating alarm strategy can improve the validity of monitoring and alarm and ensure the operation safety.
作者
周宁刚
徐剑峰
江书樵
陈泽阳
ZHOU Ninggang;XU Jianfeng;JIANG Shuqiao;CHEN Zeyang(Zhejiang Huadian Wuxijiang Hydroelectric Power Co.Ltd.,Quzhou 324000,China;Beijing HuakeTongAn Monitoring Technology Co.Ltd.,Beijing 100043,China)
出处
《水电与抽水蓄能》
2021年第4期102-106,共5页
Hydropower and Pumped Storage
关键词
水电机组
状态监测
概率统计
正态分布
波动预警
hydropower unit
condition monitoring
probability statistics
gaussian distribution
fluctuating alarm
