变电站继电保护二次回路缺陷的动态BIM识别方法

Dynamic BIM Identification Method for Secondary Circuit Defects in Substation Relay Protection

在继电保护装置更新、完善与升级后,二次回路仍然存在着较大的缺陷率,因此加大了变电站运行的风险,对此,提出变电站继电保护二次回路缺陷动态BIM识别方法。深入分析二次回路结构与缺陷类别,应用BIM技术构建变电站数据化/信息化模型,采集二次回路缺陷信息,通过小波包分析算法消除二次回路缺陷信号中的噪声,基于小波包能量谱提取二次回路缺陷特征信号向量,将缺陷特征信号向量作为输入向量训练LSSVM分类器,确定分类器最佳参数。将待识别继电保护二次回路信号输入训练好的LSSVM分类器,即可实现二次回路缺陷动态BIM的分类识别。实验数据显示,在不同实验工况背景下,提出方法的缺陷识别时延最小值为15.20 ms,缺陷识别误差最小值为0.23%,充分证实了提出方法应用性能较佳。

After the renewal,improvement and upgrading of relay protection devices,there is still a large defect rate in the secondary circuit,which increases the risk of substation operation.Therefore,the dynamic BIM identification method of substation relay protection secondary circuit defects is proposed.This paper deeply analyzes the secondary circuit structure and defect category,apply BIM technology to build the substation data/information model,collect the secondary circuit defect information,eliminate the noise in the secondary circuit defect signal through wavelet packet analysis algorithm,extract the secondary circuit defect characteristic signal vector based on wavelet packet energy spectrum,train the LSSVM classifier with the defect characteristic signal vector as the input vector,and determine the best parameters of the classifier.The secondary circuit signal of relay protection to be identified is input into the trained LSSVM classifier to realize the classification and identification of secondary circuit defect dynamic BIM.The experimental data show that under different experimental conditions,the minimum time delay of the proposed method is 15.20 ms and the minimum error of defect identification is 0.23%,which fully proves the better application performance of the proposed method.