基于最小二乘双支持向量机的配电网短路故障辨识方法研究.docx
版权申诉
19 浏览量
2023-02-23
20:32:05
上传
评论
收藏 949KB DOCX 举报
摘要
配电网短路故障时有发生，若无法及时辨识和清除故障，对电网的稳定运行和用户的生产生活都将
造成严重影响。为了对配电网短路故障类型进行准确辨识，对传统支持向量机算法进行了改进，搭
建了基于决策树和最小二乘双支持向量机的多分类辨识模型，并将其应用于配电网短路故障辨识中。
首先在短路故障发生后对采样后的电流暂态信号作希尔伯特-黄变换，其次分别提取变换后各电流
分量的能量特征构成特征向量，最终输入多分类辨识模型，实现对配电网短路故障类型的快速准确
识别，并通过仿真验证了该方法的可靠性和有效性。
Abstract
Short-circuit faults in distribution networks occur from time to time. If the faults
cannot be identified and cleared in time, there will be a serious impact on the stable
operation of the power grid and the production and life of users. In order to identify
the types of short-circuit faults in the distribution network accurately, this paper
improved the traditional support vector machine algorithm, built a multi-class
identification model based on decision trees and least squares twin support vector
machine, and applied it to the short-circuit fault identification in distribution
networks. First, Hilbert-Huang transformation was performed on the sampled
transient current signal after the occurrence of a short-circuit fault. Secondly, the
energy characteristics of each current component after the transformation were
extracted to form a feature vector, which was finally input to the multi-class
identification model to realize the rapid and accurate identification of the types of
short-circuit faults in the distribution network. The reliability and effectiveness of
the method were verified by simulation.
译
关键词
短路故障; 类型辨识; 希尔伯特-黄变换; 最小二乘双支持向量机
Keywords
short-circuit fault; type identification; Hilbert-Huang transform; least squares twin
support vector machine
译
随着我国电力系统不断发展，配电网规模日渐扩大，各类故障也时有发生。当配电网线路发生短路
故障时，系统电压大幅下降，短路点附近线路的电流急剧增加，若未及时识别故障并采取必要的保
护动作， 可能会造成线损 增大、设备烧毁 等危害，甚至造成发电机与 系统解列，导致 大面积停电。
短路故障时间越长，对电气设备和配电网系统的危害越严重。因此，快速准确地辨识、定位和处理
配电网短路故障，有助于恢复系统稳定运行、保障用户电能质量、降低故障所造成的损失
［ 1-3 ］
。短
路故障可分为单相接地短路、两相短路、两相接地短路以及三相短路 4 种类型，只有精准辨识故障
类型，才能进一步对故障进行定位和切除。
剩余10页未读，继续阅读
资源评论
罗伯特之技术屋
- 粉丝: 3652
- 资源: 1万+
下载权益
C知道特权
VIP文章
课程特权
VIP享7折，此内容立减5.97元 开通VIP
