Running status of electricity grid equipment has respect to power quality, supply reliability and other power system core indi⁃cators. This paper analyzes the traditional equipment status evaluation method, a method based on time series, adaptive neural net⁃works, unsupervised clustering methods is proposed. By the depth of mining potential operating data information, development trends of abnormal state for equipment are found, it can enhance the level of power system operation and maintenance. Study results also be ap⁃plied to specific case studies, the results show that the constructing method is reasonable and effective and can be extended to the status evaluation of transmission change.%电网设备的运行状态关系到电能质量、供电可靠性等电力系统核心指标。在分析传统设备状态测评方法的基础上，提出以时间序列、自适应神经网络、无监督聚类等方法来深度挖掘运行数据的潜在信息，发现设备向异常状态发展的趋势，以提升电力系统运维水平。同时将研究成果应用到具体案例中。结果表明，该构建方法具有一定合理性和有效性，能推广到输变电设备的状态测评中。
展开▼
