摘要
为了提高周期性势能负荷电机电力系统的稳定性,需要进行周期性势能负荷电机电力系统故障数据的自动化识别,提出基于深度学习算法的电力系统故障数据自动识别方法。建立周期性势能负荷电机电力系统故障数据的分数间隔均衡采样模型,结合对数据的结构重构输出,通过特征提取的方法实现对周期性势能负荷电机电力系统故障数据的滤波检测。通过深度学习算法,建立电力系统故障数据的特征辨识参数分析模型,根据特征提取结果,实现故障数据自动识别。测试表明,采用该方法进行电力系统故障数据识别的精度较高,自动化性能较好。
In order to improve the stability of periodic potential load motor power system,it is necessary to automatically identify the fault data of periodic potential load motor power system,and an automatic identification method of power system fault data based on deep learning algorithm is proposed.The fractional interval equalization sampling model of periodic potential energy load motor power system fault data is established.Combined with the structure reconstruction output of the data,the filtering detection of periodic potential energy load motor power system fault data is realized by feature extraction method.Through the deep learning algorithm,the feature identification parameter analysis model of power system fault data is established,and the automatic identification of fault data is realized according to the feature extraction results.The test shows that this method has high accuracy and good automation performance.
作者
张巍
张佳艺
王善立
信茜
郑丽娟
ZHANG Wei;ZHANG Jiayi;WANG Shanli;XIN Xi;ZHENG Lijuan(Hainan Power Grid Co.,Ltd.,Haikou 570203,China;Tellhow Software Co.,Ltd.,Nanchang 330200,China)
出处
《通信电源技术》
2021年第18期111-113,共3页
Telecom Power Technology
关键词
深度学习算法
电力系统
故障数据
自动识别
deep learning algorithm
power system
fault data
automatic recognition
