基于数学形态谱和人工神经网络的高压输电线接地故障类型识别方法

A Ground-Fault Type Identification Method Based on Artificial Neural Networks and Pattern Spectrum in High Voltage Transmission Lines

高压输电线接地故障的正确、快速识别是处理故障的前提之一。该文提出了一种基于数学形态谱和人工神经网络识别高压输电线路接地故障类型的新方法。数学形态学颗粒分析是一种用来处理图像的粒度和形状特征的图像处理工具。该方法通过对故障电流进行相模变换后,用数学形态学颗粒分析方法提取序电流分量的形态谱,并作为神经网络的输入,实现对接地故障类型的识别。仿真表明,该方法具有较高是识别率。

The accurate quick ground-fault type identification in high-voltage transmission lines is one of the prerequisites to deal with faults. In this paper, a method based on mathematical morphology and artificial neural networks is proposed to identify the ground-fault of high-voltage transmission lines. Mathematical analysis of the particle morphology is a very effective tool for image processing which is mainly used to deal with the image size and shape features. This method, which converts the current of ground-fault phase sequence to mode sequence first, uses mathematical analysis of the morphology of particles to extract the spectrum shape of sequence current component. The artificial neural network is used to process the spectrum, which implements the identification of ground-fault type. The simulation shows that the method achieves a high recognition rate.