基于自学习的电网设备故障自动辩识方法研究与应用

Research on automatic identification of abnormal events in monitoring substation

电网设备故障自动辨识是一项可实现变电站无人值守、保障变电站安全、稳定运行的基础性支撑技术。文中针对传统电网设备故障识别准确率低的问题,提出了一种结合SIFT和深度学习算法的自动辨识控制系统设计方法。通过集中监控获取变电站的实时视频、图像信息,利用SIFT对数据预处理并结合深度学习算法,实现了电网设备故障的自动辨别。实验证明,基于该方法所设计的自动辨别系统提高了电网设备故障辨识准确率。

Automatic identification of substation abnormal events is a basic supporting technology to realize unattended substation and ensure safe and stable operation of substation.Combining SIFT and deep learning algorithm,this paper studies an automatic identification system for monitoring substation abnormal events.The real-time video and image information of substation is obtained by centralized monitoring,and the data is preprocessed by SIFT.Through machine learning algorithm,the automatic identification of abnormal events in substations is realized,and the automation degree of substation monitoring and identification is improved.Experiments show that this method can provide some technical support for unattended substation monitoring,and has a certain practical value.