基于关联规则的高压换流站典型直流运维事件诊断策略

Typical DC O&M Event Diagnosis Strategy of High Voltage Converter Station Based on Correlation Rules

针对高压换流站直流系统在状态转变或者发生故障时,其关于换流站顺序事件记录的系统在瞬时会产生大量该事件。而这种隐形的故障,很难用人工的方法观察到其相关的故障特征。因为,针对该问题本文基于关联规则来进行对发生的事件或者故障来进行判断。首先,针对该事件的典型特征种类进行分析并筛选,之后,通过关联规则算法来对该事件可能相关或者相似的故障类型进行挖掘。依据上述挖掘产生的结果来判断。并通过具体算例进行分析验证。验证结果表明本文方法具有较好的鲁棒性。

For the high-voltage converter station DC system in the state transition or failure,the system of the converter station sequence event record will generate a large number of these events instantaneously.This invisible fault makes it difficult to observe its related fault characteristics by manual methods.Because,for this problem,this article is based on the correlation rules to determine the occurrence of events or failures.First,the typical characteristics of the event are analyzed and screened,and then the correlation rule algorithm is used to mine the types of failures that may be related or similar to the event.Judge based on the results of the above excavations.It is analyzed and verified by specific studies.The results show that the proposed method has good robustness.