摘要
变电站中高压开关柜的安全运行直接影响着供电可靠性,以状态维修为目的的状态评价是评估其运行状态的有效手段。本文选取暂态对地电压幅值、超声波幅值、暂态对地电压脉冲数、开关柜基本信息评分值、预先测试检定数据评分值的数据样本作为评估的5个状态量,以欧式距离作为约束依据隶属度最大原则进行聚类,在此基础上提出了一种基于模糊聚类算法的开关柜评价方法。以实际变电站开关柜带电检测数据为例对开关柜状态进行评估,验证了该方法的有效性。
The safe operation of HV switch cabinet in substation directly affects the reliability of power supply.The state evaluation for condition maintenance is an effective means to evaluate its operation status.The data samples of transient grounding voltage amplitude,ultrasonic amplitude,transient grounding voltage pulse number,switch cabinet basic information score value and pre-test verification data score value are selected as the five state variables for the evaluation.Clustering is based on the principle of maximum membership degree with Euclidean distance as constraint.On this basis,a switch cabinet evaluation method based on the fuzzy clustering algorithm is proposed.Taking the live detection data of switch cabinet in substation as an example,the status of switch cabinet is evaluated,and the effectiveness of this method is verified.
引文
[1]潘长明,刘刚,熊炬,等.高压开关柜绝缘事故的分析及防范措施[J].高压电器,2011,47(7):90-93.
[2]谢静,束洪春,王科,等.基于FCM算法的高压开关柜局部放电状态评价方法研究[J].高压电器,2015(10):82-90.
[3]谢静,束洪春,王科,等.基于模糊分层理论的高压开关柜状态评估算法[J].高电压技术,2014,40(10):3186-3192.