融合多源数据的架空输电线路运行状态评估方法

Condition assessment method of overhead transmission lines based on multi-source data

针对当前输电线路的状态评估指标量互相孤立及评估方法准确率较低的不足,提出了一种基于信息融合思想的架空输电线路运行状态评价方法。首先选取线路试验和巡视信息、运维检修记录、电网运行参量等多源数据作为指标量,采取隶属度函数表达指标量的分布状态,并以分布状态为数据标签,建立基于长短时记忆网络的线路运行状态综合评价模型。利用数据标签训练该深层网络,以建立指标参量与运行状态之间的映射关系。案例分析表明,所提方法的评估准确率高达95.5%,较高于传统浅层方法的评估准确率,可为架空输电线路运维提供理论指导。

Due to the isolation of assessment indexes and the low accuracy of assessment methods, this paper proposes a condition assessment method of overhead transmission lines based on information fusion. Firstly, the multi-source data from line test and inspection, operation and maintenance records, and line operation parameters are selected as assessment indexes. The membership functions are utilized to express the distribution status and data label of indexes. Thus, a comprehensive assessment model of transmission line operating condition based on long short-term memory network is established. The deep network is trained by the data labels, thus the mapping relationship between the indexes and condition grades is establish. The case shows that the assessment accuracy of the proposed method is as high as 95.5%, which is higher than the assessment accuracy of traditional shallow methods. This study can provide theoretical guidance for the operation and maintenance of overhead transmission lines.