基于规则网的高压输电线路故障诊断与保护动作性能评价

A Rule-Network Based Approach for Fault Diagnosis of HV Transmission Lines and Protection Action Performance Evaluation

对于多重或复杂故障以及存在警报信息丢失或畸变等复杂情况,如何准确诊断所发生的故障和评价保护装置的动作性能仍是值得研究的问题。在此背景下,提出一种基于规则网利用故障录波信息进行高压输电线路故障诊断和评价保护与断路器动作性能的新方法。首先,对故障录波信息进行预处理,利用小波变换技术从电气模拟量信息中提取线路故障特征量,得到电流突变事件和保护动作事件序列。在此基础上,建立计及警报信息时序特性的高压线路故障诊断规则网络模型,并确定故障假说集与期望的保护和断路器状态,并利用它们之间存在的因果与时序关系确定故障原因以及对保护和断路器动作性能进行评价;之后,对故障录波做进一步分析以判别故障类型、故障性质和故障位置;最后,用实际电力系统故障案例对所提出的方法进行说明。

For the situations of complicated and multiple faults with missing and/or false alarms, it is still a challenging task to develop an accurate approach for fault diagnosis of high-voltage （ HV ） lines and operating behavior evaluation of protective devices. Given this background, this paper presents a rule-network based approach for this purpose by employing fault recorder data. Firstly, the fault recorder data are pre-processed by using the well-developed wavelet transform to extract the fault characteristics of HV lines from analogue electrical information, and the events of sudden current change and protection action are obtained. On this basis, we construct a rule-network based model for fault diagnosis of HV lines with considering the temporal characteristics of alarm messages, determine the sets of fault hypotheses as well as the expected states of protective relays and circuit breaker, and determine the fault cause and evaluate the performance of protection and circuit breaker with using the causal and temporal relationships between them. In addition, we further analyze the fault recorder to determine the type, characteristics and location of fault. Finally, we adopt actual power system fault cases to demonstrate the proposed method.