基于Bayes和Bootstrap方法的智能电表可靠性评估  

Reliability Evaluation of Smart Meters Based on Bayes and Bootstrap Methods

作　　者：李若茜 肖霞[1] 梅能 柯航 LI Ruoqian;XIAO Xia;MEI Neng;KE Hang(School of Electrical and Electronic Engineering,Huazhong University of Science and Technology,Wuhan 430074,China)

机构地区：[1]华中科技大学电气与电子工程学院,武汉430074

出　　处：《南方电网技术》2022年第3期76-81,共6页Southern Power System Technology

基　　金：国家自然科学基金资助项目(51821005)。

摘　　要：针对智能电表现场数据不能涵盖其整个生命周期、加速寿命试验应力同实际运行环境存在差异使得仅以单一数据源为依据的可靠性评估结果不够准确的问题,提出了结合Bayes和Bootstrap方法的智能电表可靠性评估方法。该方法采用Bootstrap方法处理现场数据,得到智能电表可靠性模型参数的离散分布,以该离散分布作为先验信息,采用Bayes方法结合加速寿命试验数据得到融合两种信息后的参数估计值,实现智能电表的可靠性评估。实例结果表明,该方法得到的智能电表可靠性模型在前半段贴近现场实际情况,在后半段有涵盖全生命周期的试验数据作为支撑,可用于智能电表整个生命周期的可靠性评估。Due to the reliability evaluation result based on a single data source is not accurate enough caused by the field data of the smart meters not covering its entire life cycle and the difference between the stress of accelerated life test and the actual operating environment,a reliability assessment method of smart meters combining Bayes and Bootstrap methods is proposed.The proposed method uses Bootstrap method to process the field data to obtain the discrete distribution of the reliability model parameters of the smart meters,then adopts the Bayes method to fuse the discrete distribution,as prior information,and the accelerated life test data to obtain the parameter estimates,and realizes the reliability assessment of smart meters.The result of the example shows that the reliability model of the smart meters obtained by the proposed method approximates the actual field situation in the first half and is supported by test data in the second half,which can be used for the reliability assessment of the entire life cycle of the smart meters.

关 键 词：智能电表 可靠性评估 BOOTSTRAP BAYES 数据融合 

分 类 号：TM993[电气工程—电力电子与电力传动]