摘要
暴雨、台风等强对流天气导致的故障会对电力系统的安全稳定运行造成极大威胁,因此需要建立有效的故障预警模型。利用深度融合神经网络与多源数据相融合的方法建立极端天气的时空变化预测模型,同时结合输电杆塔的故障分析建立电力设备的故障预测。实验表明:神经网络最终的MAE为0.5,RMSE为0.2,均低于其他方法;在滑坡场景中,4~8号杆塔的损毁概率均超过了0.5,其中,7号杆塔的损毁概率最高,为0.79;杆塔的故障率随着降雨量的增加逐渐上升。结果表明,该预测模型能够对极端天气下电力设备的故障进行预测,且具有较好的效果。
The normal operation of power equipment is affected by extreme weather such as rainstorm and typhoon,hence,it is necessary to establish an effective fault warning model.The deep fusion neural network and multi-source data fusion method are used to establish a spatiotemporal prediction model for extreme weather.A fault analysis for transmission tower is used to establish this method.The final MAE and RMSE of neural network are 0.5 and 0.2,respectively,which are both lower than other methods.In the landslide scenario,the damage probability of Towers 4 to 8 exceeded 0.5,and Tower 7 has the highest value of 0.79.The failure rate of tower gradually increases when rainfall increases.The proposed prediction model can predict the faults of power equipment under extreme weather and has better results.
作者
侯艳权
王义春
李岑
佘燕飞
修唯
掌旭
HOU Yanquan;WANG Yichun;LI Cen;SHE Yanfei;XIU Wei;ZHANG Xu(Qitaihe Power Supply Company of State Grid Heilongjiang Electric Power Co.,Ltd.,Qitaihe 154600,China;Shenyang Jiayue Electric Power Technology Co.,Ltd.,Shenyang 110136,China)
出处
《微型电脑应用》
2024年第8期73-76,共4页
Microcomputer Applications
基金
基于“电力+气象”智慧气象服务平台建设的研究与应用(522424230004)。
