摘要
针对增量配电网调度策略弱、效益低的问题,提出了一种高分辨率滚动周期调度策略和一种基于深度学习的长短周期比较多步电价预测模型.以某地区2012~2015年市场电价数据及各项供需输送关系数据为基础,结合滚动周期调度策略和空洞卷积神经网络建立增量配电网储能系统调度模型,对储能系统调度过程中电价走势及未来数据进行预测.结果表明,相比于目前常用的研究方法,所提出模型在经济效益上提高了9.1%,并具有更优秀的鲁棒性和效率.
Aiming at the problem of weaker scheduling strategy and lower economic benefits of incremental distribution network,a high-resolution rolling period scheduling strategy and a multi-step electricity price forecast model with long-short term comparison based on deep learning were proposed.According to the history data of market electricity price and the supply-demand transport relationship data in a certain area from 2012 to 2015,a scheduling model for energy storage system in incremental distribution network was established,in combination with rolling period scheduling strategy and dilated convolutional neural network,and the electricity price trend during the scheduling process of energy storage system and the future data were forecasted.The results show that the as-proposed model brings about 9.1%increase of economic benefits and exhibits better robustness and efficiency compared with the commonly used methods.
作者
宋坤
武志锴
南哲
张明慧
李大东
SONG Kun;WU Zhi-kai;NAN Zhe;ZHANG Ming-hui;LI Da-dong(Software College,Northeastern University,Shenyang 110169,China;Economic Research Institute,State Grid Liaoning Electric Power Supply Co.Ltd.,Shenyang 110015,China;Financial Assets Department,State Grid Liaoning Electric Power Supply Co.Ltd.,Shenyang 110015,China;Financial Assets Department,State Grid Liaoning Bidding Co.Ltd.,Shenyang 110051,China)
出处
《沈阳工业大学学报》
CAS
北大核心
2021年第5期493-499,共7页
Journal of Shenyang University of Technology
基金
辽宁省自然科学基金项目(2019-MS-112)
中央高校基本科研业务专项资金项目(N2017001)
国家电网科技项目(SGLNJY00ZLJS2000091).
关键词
电价预测
价格峰值检测
滚动周期策略
储能系统
深度学习
运行调度
电力调度
策略优化
electricity price forecast
price peak detection
rolling period strategy
energy storage system
deep learning
operation scheduling
power scheduling
strategy optimization
作者简介
宋坤(1978-),男,辽宁丹东人,高级工程师,硕士,主要从事能源电力开发技术、电力系统规划与设计等方面的研究.
