基于深度学习的110 kV输配电线路运行故障检测研究

Research on Fault Detection of 110 kV Transmission and Distribution LinesBased on Deep Learning

为优化110 kV输配电线路运行故障检测效果,提高故障检测率,保证输配电线路安全可靠运行,引入深度学习原理,开展了基于深度学习的110 kV输配电线路运行故障检测研究。首先,建立110 kV输配电线路数学模型,更好地获取线路运行工况及电力参数。其次,采用有效值法计算线路各区段零序电流有效值与所有区段零序电流有效值平均值之比,根据计算结果,判断线路运行故障区段。在此基础上,利用深度学习原理构建深度学习网络结构,提取故障区段的故障分量,计算线路故障距离,进而在故障区段内根据故障距离,确定故障点,实现故障检测定位目标。结果表明,应用提出的检测方法后,6条110 kV输配电线路运行的故障检测率均达到98%以上,能更加准确地检测输配电线路运行中潜在的故障隐患。

In order to optimize the fault detection effect of 110 kV transmission and distribution lines,improve the fault detection rate,and ensure the safe and reliable operation of transmission and distribution lines,the principle of deep learning was introduced to carry out research on 110 kV transmission and distribution line operation fault detection based on deep learning.Firstly,establish a mathematical model for the 110 kV transmission and distribution line to better obtain the operating conditions and power parameters of the line.Secondly,the effective value method is used to calculate the ratio of the effective value of zero sequence current in each section of the line to the average effective value of zero sequence current in all sections.Based on the calculation results,the faulty section of the line operation is determined.On this basis,using the principle of deep learning,a deep learning network structure is constructed to extract the fault components of the fault section,calculate the fault distance of the line,and then determine the fault point within the fault section based on the fault distance,achieving the goal of fault detection and positioning.The results show that after the application of the proposed detection method,the fault detection rate of 6110 kV transmission and distribution lines has reached over 98%,which can more accurately detect potential fault hazards in the operation of transmission and distribution lines.