摘要
准确地拟合负荷数据是表征负荷特性的关键环节。文章以电力网建设为背景,提出一种改进变分模态分解(Improved Variational Mode Decomposition,IVMD)和长短期记忆网络(Long Short-Term Memory,LSTM)结合的电网中长期典型时段负荷曲线拟合方法。该方法首先将曲率作为量化指标,减小常规VMD方法中K值的不确定性;其次,考虑高渗透率灵活负荷并网条件,精细化拟合特殊时段电网负荷,给出IVMD-LSTM的自适应电网负荷降噪及曲线拟合方法,提高负荷曲线拟合精度和效率;最后,通过实际算例证明了所提方法可有效减小电网典型时段负荷曲线拟合误差,辅助弃风消纳。
Accurate load data fitting is the key to characterize the load characteristics. Most of the existing methods are difficult to meet the grid scheduling requirements in some typical time period.To tackle this disadvantage, an typical load curve fitting of grids is proposed with improved variational mode decomposition(IVMD) and long short-term memory(LSTM) based on the inspection method of ubiquitous power internet of things construction. Firstly, the load curvature is proposed as the quantitative index for the first time. And the uncertainty of the K value is decreased effectively than the conventional VMD method which depends on the experience.Secondly, for the case of high-percentage grid-connected flexible load, the grid load during typical time period is fit in detail. And the IVMD-LSTM adaptive grid load noise reduction and curve fitting method is given to improve the accuracy and efficiency of load curve fitting. Finally,an actual example results show that the proposed method reduces the fitting error of the load curve fitting in the typical period of and the scheduling cost of the grid load effectively.
作者
胡博
潘霄
葛维春
崔嘉
杨俊友
徐熙林
Hu Bo;Pan Xiao;Ge Weichun;Cui Jia;Yang Junyou;Xu Xilin(StateGrid Liaoning Electric Power Co.,Ltd.,Shenyang 110000,China;Economic and Technology Research Institute State Grid Liaoning Electric Power Co,Ltd,Shenyang 110015,China;The School of Electrical Engineering Shenyang University of Technology,Shenyang 110870,China)
出处
《可再生能源》
CAS
北大核心
2020年第3期366-372,共7页
Renewable Energy Resources
基金
国家电网项目(SGTYHT17JS199)
辽宁省教育厅科学技术项目(201634091)
辽宁省自然科学基金(20170520318)
