摘要
光伏发电具有随机性、间歇性等特点,其大规模接入配电网给电网安全、经济运行带来挑战。储能是解决高渗透率光伏输出功率波动的有效手段。为此,提出一种结合相邻时段实时预测误差和极限学习机的光伏功率区间预测方法;提出一种分布式储能的区间离散SOC均衡控制策略。在此基础上,构建了包含储能运行成本、系统网损和电压偏差的多目标区间优化运行模型。应用区间泰勒展开的仿射算法将不确定区间潮流问题转化为确定性代数方程组实现潮流计算,同时依据电压–功率和网损–功率灵敏度矩阵重新表达目标函数,采用线性规划法求解。最后,基于改进的IEEE33节点配电网络进行了算例仿真,仿真结果表明,所提光伏功率区间预测方法可以显著提升预测精度,所提储能控制策略能够有效降低网损和改善电压分布。
Large-scale connection of photovoltaic power generation to the distribution network poses challenges to the safety and economic operation of the power grid owing to its random and intermittent characteristics. Energy storage is an effective method for solving the high-permeability photovoltaic output power fluctuations. In this study, a photovoltaic power interval prediction method combining error feedback and extreme learning machine is proposed. Furthermore, an interval discrete state of charge equalization control strategy for distributed energy storage is proposed. A multi-objective interval optimization operation model is constructed, including the energy storage operating costs, system network loss, and voltage deviation. The affine algorithm of interval Taylor expansion is used to transform the uncertain interval power flow problem into deterministic algebraic equations to calculate the power flow. Simultaneously, the objective function is expressed by voltage-power and network loss-power sensitivity matrix and solved using the linear programming method. Finally, a simulation is performed based on the improved IEEE 33 bus distribution network. Simulation results show that the proposed photovoltaic power interval prediction method can significantly improve the prediction accuracy, and the proposed energy storage control strategy can effectively reduce network loss and improve voltage distribution.
作者
袁晓冬
郁正纲
张宸宇
伏祥运
李建林
YUAN Xiaodong;YU Zhenggang;ZHANG Chenyu;FU Xiangyun;LI Jianlin(State Grid Jiangsu Electric Power Company Research Institute,Nanjing 211103,Jiangsu Province,China;Lianyungang Power Supply Company,State Grid Jiangsu Electric Power Company,Nanjing 222004,Jiangsu Province,China;China Electric Power Research Institute,Haidian District,Beijing 100192,China)
出处
《全球能源互联网》
2019年第6期598-607,共10页
Journal of Global Energy Interconnection
基金
国家电网公司总部科技项目(5210D0180006).
关键词
光伏
储能
区间预测
SOC均衡
仿射算法
灵敏度矩阵
photovoltaic
energy storage
interval prediction
SOC equilibrium
affine algorithm
sensitivity matrix
作者简介
通信作者:袁晓冬(1979),男,研究员级高工,研究方向为电能质量、新能源发电及并网、电力储能应用等,E-mail:lannyyuan@hotmail.com,1838658@qq.com;郁正纲(1972),男,研究员级高工,研究方向为电力系统运行与分析、状态监测、可再生能源发电并网等;张宸宇(1989),男,高级工程师,研究方向为直流配电网、电能质量治理等。
