绿色电力理念下基于改进YOLOX算法的变压器节能与可靠性研究分析  

Research and Analysis on Energy Conservation and Reliability of Transformers Based on Improved YOLOX Algorithm under the Concept of Green Power

作　　者：王宏 宋禹飞 窦如婷 王昕 王庆红 WANG Hong;SONG Yufei;DOU Ruting;WANG Xin;WANG Qinghong(Southern Power Grid Research Institute Co.,Ltd.,Guangzhou 510663,China;China Southern Power Grid,Guangzhou 510663,China)

机构地区：[1]南方电网科学研究院有限责任公司,广州510663 [2]中国南方电网有限责任公司,广州510663

出　　处：《自动化与仪器仪表》2024年第11期208-212,共5页Automation & Instrumentation

摘　　要：在绿色电力理念的背景下,研究为了解决传统的变压器监测不仅效率低下,而且能耗较高的问题,研究首先对尤洛克斯算法进行了一系列改进,并融合强化学习算法设计出一种新的变压器节能模型。实验结果表明,48座变电站在应用该方案后的单位时间能耗得到了明显的降低,平均值约为4210 kJ,相比应用前下降了8.76%。以上结果表明,研究提出的变压器节能优化模型具有高可靠性,低能耗等优点,为绿色经济发展提供了新方案。In the context of the concept of green power,in order to solve the problem of low efficiency and high energy consumption in traditional transformer monitoring,the study first made a series of improvements to the Eulox algorithm and integrated reinforcement learning algorithms to design a new transformer energy-saving model.The experimental results show that the energy consumption per unit time of 48 substations has been significantly reduced after applying this scheme,with an average value of about 4 210 KJ,a decrease of 8.76% compared to before application.In summary,the energy-saving optimization model for transformers proposed in the study has advantages such as high reliability and low energy consumption,providing a new solution for the development of green economy.

关 键 词：YOLOX 强化学习 变压器 绿色电力 

分 类 号：TM42[电气工程—电器]