基于深度学习的多能互补供需自适应调度方法  

Multi-energy complementary supply and demand adaptive scheduling method based on deep learning

作　　者：刘志宏 旦增巴桑 夏强 廖晓群[2] LIU Zhihong;DANZENG Basang;XIA Qiang;LIAO Xiaoqun(Economic and Technological Research Institute of State Grid Tibet Electric Power Co.,Ltd.,Lhasa 850000,China;Center of Information and Network,Xi’an University of Science and Technology,Xi’an 710054,China)

机构地区：[1]国网西藏电力有限公司经济技术研究院,西藏拉萨850000 [2]西安科技大学网络中心,陕西西安710054

出　　处：《电子设计工程》2020年第24期108-111,116,共5页Electronic Design Engineering

基　　金：国网西藏电力有限公司科学技术项目(SGXZJY00JHJS2000008)。

摘　　要：为平衡电网供需电子间的互补调度关系,提出基于深度学习的多能互补供需自适应调度方法。分级连接电子识别控制器与多能监测模块,获取精确的学习空间复杂度数值,完成基于深度学习的多能电子识别处理。在此基础上,建立自适应调度框架,借助基层组织内电子供需结构,计算互补偏离度的实值参量,完成新型多能互补供需自适应调度方法。实验结果表明,与传统非动态调度模型相比,所提方法将单位时间内供需输出电压提升至440 V,且提高了邻位电子间的调频速率,平衡了电网供需电子间的互补调度关系。In order to balance the complementary scheduling relationship between the supply and demand of power grid,an adaptive scheduling method based on deep learning is proposed.Hierarchical connection of electronic recognition controller and multi-function monitoring module can obtain accurate learning space complexity value and complete multi-function electronic recognition processing based on deep learning.On this basis,an adaptive scheduling framework is established.With the aid of the electronic supply and demand structure in the grass-roots organizations,the real value parameters of the complementary deviation degree are calculated,and a new adaptive scheduling method of multi energy complementary supply and demand is completed.The experimental results show that compared with the traditional non dynamic scheduling model,the proposed method can increase the output voltage of supply and demand to 440 V per unit time,and improve the frequency modulation rate between adjacent electrons,balancing the complementary scheduling relationship between supply and demand electrons.

关 键 词：深度学习 多能互补供需 自适应调度 识别控制器 电子供需 互补偏离度 

分 类 号：TM929[电气工程—电力电子与电力传动]