基于深度学习的变电站硬压板状态检测与识别算法

针对变电站的一键顺控停、送电操作不当将导致投退操作的问题,提出了一种基于深度学习的变电站硬压板状态检测与识别算法.使用一个共享网络提取图像特征,基于多任务学习方法建立3个分支联合解决硬压板位置检测、投切状态检测和标识检测这3个任务;采集标注了8 000张硬压板图片数据用于训练和测试.结果表明,所提出的方法能够在提升硬压板状态识别精度的同时,也提升一键顺控操作的安全性.

State detection and recognition algorithm for hard platens of substation based on deep learning

Aiming at the problem that the one-key sequential control of stopping and improper power transmission operation of substation can lead to throwback and retreat operation, a state detection and recognition algorithm for hard platens of substation based on deep learning was proposed. A shared network was used to extract image features, and three branches were established according to a multi-task learning method to jointly finish three tasks, i. e. position detection, switching state detection and identity detection of hard platens. In addition, 8 000 hard platens were collected and labeled for training and testing. The results show that the as-proposed method can improve the accuracy of state recognition of hard platens, and it also improves the safety of one-key sequential operation at the same time.