电器与能效管理技术 ›› 2023, Vol. 0 ›› Issue (3): 62-68.doi: 10.16628/j.cnki.2095-8188.2023.03.010

• 检测与试验 • 上一篇    下一篇

基于神经网络算法的开关柜温荷研究与应用

徐卫东, 何文志, 廖肇毅, 刘勤锋   

  1. 广东电网有限责任公司 东莞供电局, 广东 东莞 523000
  • 收稿日期:2022-07-09 出版日期:2023-03-30 发布日期:2023-04-11
  • 作者简介:徐卫东(1992—)男,工程师,主要从事电力系统在线监测技术、故障诊断的研究。|何文志(1985—),男,高级工程师,主要从事电力电子与电力传动研究。|廖肇毅(1973—),男,高级工程师,主要从事电气试验及状态监测研究。

Research and Application of Switchgear Temperature Load Based on Neural Network Algorithm

XU Weidong, HE Wenzhi, LIAO Zhaoyi, LIU Qinfeng   

  1. Dongguan Power Supply Bureau,Guangdong Power Grid Co.,Ltd., Dongguan 523000, China
  • Received:2022-07-09 Online:2023-03-30 Published:2023-04-11

摘要:

大电流开关柜温度与负荷电流异常变化可能会导致设备缺陷进而影响正常运行。为了对开关柜的温度实现有效监测,采用Ansoft电-热学仿真模块分析了开关柜内部关键部件温度场变化情况,获取了不同环境下温度变化的数据,掌握了温度传感器与发热点距离位置的关系。最后,通过试验获取不同电流下温荷曲线,利用神经网络预测方法与现场试验相结合的方式验证了研究结果的准确性。试验结果能够为变电站现场运维人员的温度传感器安装方案提供指导。

关键词: 大电流开关柜, 温度传感器, 神经网络仿真, 温升试验

Abstract:

Abnormal changes in temperature and load current of high-current switchgear may lead to equipment defects and affect normal operation.In order to effectively monitor the temperature of the switchgear,the Ansoft electro-thermal simulation module is used to analyze the temperature field changes of key components in the switchgear.The data of temperature changes under different environmental conditions is obtained.The relationship between the temperature sensor and the distance and position of the hot spot is mastered.Finally,the temperature load curve under different current is obtained through the test.The accuracy of the research results is verified by combining the neural network prediction method with the field test.The results can provide the guidance for the on-site operation and maintenance personnel of the substation to install the temperature sensor.

Key words: high-current switchgear, temperature sensors, neural network simulation, temperature rise experiment

中图分类号: