摘要
本文采用概率分析手段依据中心极限定理对三峡水电站机组在线监测振摆长期数据进行统计,发现机组振摆混频幅值服从N(μ,σ2)的正态分布规律。根据混频幅值分布规律提出了报警阈值空间的3σ原则,制定了合理的报警策略。以定水头变负荷为例对报警阈值空间进行详细讨论。研究表明本文所研究的报警策略可以在实践基础上进一步发展,从而对机组长期潜在性危险因素进行跟踪。这一报警方式较ISO7919.5和ISO10816.5规定的单一报警值具有非常大的进步。
This study did a statistical analysis on the long-term series of online maintenance vibration data for the Three Gorges hydropower station by applying the central limit theorem and revealed a normal distribution for the peak-to-peak value of generator vibrations. Based on this distribution we put forwards an alarm threshold principle, namely 3σ principle, and a corresponding alarm strategy. In a case study of constant water head and varied load, selection of alarm threshold was analyzed in details. This alarm strategy is much improved relative to those in ISO7919.5 and ISO10816.5, and it could be further developed for application to monitoring potential dangers of hydraulic turbine generators.
出处
《水力发电学报》
EI
CSCD
北大核心
2013年第5期269-272,293,共5页
Journal of Hydroelectric Engineering
基金
中国水利水电科学研究院青年科技专项(计集1205)
关键词
水轮发电机组
中心极限定理
统计特性
状态监测
报警阈值
hydraulic turbine generator
central limit theorem
statistical characteristic
status monitoring
alarm threshold
