基于威布尔分布的电气设备缺陷率模型参数分析  

Parameter Analysis of Defect Rate of Electrical Equipment Based on Weibull Distribution

作　　者：赵建利[1] 孙祎[1] 柴林杰 李培栋[3] 路艳巧[1] 

机构地区：[1]国网河北省电力公司电力科学研究院,石家庄050021 [2]国网河北省电力公司经济技术研究院,石家庄050021 [3]国网北京经济技术研究院,北京100052

出　　处：《河北电力技术》2015年第1期5-8,38,共5页Hebei Electric Power

摘　　要：针对传统指数模型和多项式模型曲线拟合存在的缺点,引入威布尔模型描述电气设备缺陷率发展的2个阶段,应用Levenberg-Marquardt法对模型参数进行求解,确定更为科学可靠的缺陷率突增期,并通过主网电气设备缺陷率统计数据进行算例分析,对3种拟合方法进行对比,验证了威布尔模型拟合曲线的有效性,为技改资金的分配以及运维决策的制定提供科学依据。For the disadvantages of exponential model and polynomial model, the paper uses Weibull distribution to describe the two defect progress stages of electrical equipment,and applies Levenberg-Marquardt method for model parameters solution, so as to determine the uprush period of defect rate more scientifically and reliably.At last, an instance of statistical data of defect rate of electrical equipment of major power network is analyzed, and three polynomial methods are compared to validate the feasibility, and to take reference for capital distribution and strategies of operation.

关 键 词：威布尔分布 电气设备 缺陷率 参数分析 

分 类 号：TM732[电气工程—电力系统及自动化]