基于深度学习的变压器负载能力评估与应用  

Evaluation and Application of Transformer Overload Capacity Based on Big Data

作　　者：王兴勋 王月香 倪康婷 李云鹏 WANG Xingxun;WANG Yuexiang;NI Kangting;LI Yunpeng(Beijing Guowangfuda Technology Development Co.,Ltd.,Application Department of Power Internet of Things Center,Beijing 100070,China)

机构地区：[1]北京国网富达科技发展有限责任公司电力物联网中心应用部,北京100070

出　　处：《电工技术》2022年第15期42-45,55,共5页Electric Engineering

摘　　要：在电力系统中,变电容量的不同主要是受变压器系统负载能力的直接影响。在电力变压器正常状态或事故状况下过载时,就需电力变压器不因过载而损坏,并确保有充分时间把过载运行限制在安全稳定范围内,所以针对电力变压器负载能力的预先评测与预警,对于大电网系统安全稳定运行至关重要。为此,基于大数据人工智能技术,采用温度计算模型,考虑多参量影响,融合电网运行数据,对变压器过载能力进行预测评估研究,并通过工程实践证明了该方法的有效性,在为保障变压器稳定运行提供依据的同时满足了实际运行需要。In the power system,the difference of transformer capacity is mainly directly affected by the load capacity of the transformer system.When the transformer operates under overload under normal or accident conditions,the transformer is required not to be damaged due to overload,so as to ensure sufficient time to control the overload within the safe and stable range.Therefore,the pre-evaluation and early warning of the load capacity of power transformers are very important for the safe and stable operation of large power grid systems.For this reason,based on artificial intelligence technology of large data,using temperature calculation model,considering the influence of multiple parameters,fusing the operation data of power network,the transformer overload capacity is predicted and evaluated.The validity of this method is proved by engineering practice,which provides a basis for ensuring the stable operation of transformers and meets the actual operation needs.

关 键 词：变压器 负载能力评估 电网运行 

分 类 号：TM41[电气工程—电器]