During the operation of the distribution network,excessive corona discharge leads to gap breakdown,resulting in excessive ionization of the output power signal of the switchgear,which reduces operational stability.Therefore,a deep learning based corona discharge detection method for distribution network switchgear is proposed.Using a deep learning recognition model related to corona signals,a monotonic non increasing function is defined,and based on this,the signal sampling value is calculated to achieve sampling of corona signals from distribution network switchgear under deep learning.Determine the expression of the corona signal string,and complete real-time discharge detection based on the solution results of the switchgear pulse.The experimental results show that the proposed method can control the maximum value of corona discharge below 100 pC,and can avoid excessive ionization of the output power signal of the switchgear.
关键词
深度学习/配电网开关柜/电晕放电/单调非递增函数/放电检测
Key words
deep learning/distribution network switchgear/corona discharge/monotonic non increasing function/discharge detection
维普
万方数据