基于动态聚类的负荷管理终端故障定位方法研究

Study on Load Management Terminal Fault Location Method Based on Dynamic Clustering

为了提高计量自动化终端故障定位准确率,提出了一种数值统计法和数学模型算法相结合的基于动态聚类的负荷管理终端故障定位的新方法。该方法对终端采集的电能量数据和档案信息采用动态聚类分析归类,形成负荷管理终端故障规则库,筛选出故障信息点,从而实现实时的数据监测和故障诊断及定位。为验证新方法的有效性和准确性,分别采用新方法和数值统计法对某测试对象的档案数据和电能量数据进行计算。计算结果表明:新方法诊断故障种类的准确率明显高于数值统计法。所提出的新方法为负荷管理终端故障运维提供一种技术手段。

In order to increase the accuracy rate of fault location of metrology automation terminal, a new fault location method of load management terminal based on dynamic clustering which combines numerical statistics method and mathematical model algorithm is proposed. Dynamic clustering is used to analyze and classify both electric power energy data collected by terminals and archives information in the method, and fault rule database of load management terminal is formed screen fault information point so that the real- time data monitoring and fault diagnosis and location can he realized. In order to verify the effectiveness and correctness of the new method, archives data and electric power energy data of a test object are calculated respectively by the new method and numerical statistics method. Calculation results show that the accuracy rate of fault classification diagnosed by the new method is apparently higher than numerical statistics method. The new method proposed provides a technical means for the operation and maintenance of load management terminal fault.