论文部分内容阅读
针对目前高压隔离开关的发热温度检测没有可靠在线监测设备和技术手段,导致难以对过热缺陷进行可靠预警的问题。本文基于RBF神经网络,综合考虑了影响高压隔离开关发热的负荷电流率、污秽等级和环境温度3个因素,建立了一个隔离开关过热故障的预警模型,通过对现场数据样本测试发现:模型对发热状态评估的总体正确率达到了94.44%,且做到了对过热缺陷100%的预警。
There is no reliable on-line monitoring equipment and technical means for detecting the heating temperature of the high-voltage disconnector at present, which makes it difficult to reliably warn of overheating defects. Based on the RBF neural network, an early warning model of the overheat fault of the switchgear is established by comprehensively considering the load current rate, the contamination level and the ambient temperature which affect the heating of the HV disconnector. By testing the field data samples, The overall correctness of the state assessment reached 94.44%, and made an early warning of overheating defects 100%.