摘要
针对传统动态概率潮流(dynamic probability power flow,DPPF)计算结果存在的保守性问题,提出一种计及风速预测误差时空相关性的DPPF计算方法。首先,采用自相关系数平稳过程描述输入变量的预测误差过程,利用非参数核密度估计直接根据预测误差历史数据拟合得到预测误差分布;其次,基于等概率变换理论和Nataf变换技术得到具有时空相关性的误差样本。最后,通过基于拉丁超立方采样的蒙特卡罗模拟法进行DPPF计算,得到节点电压幅值和支路潮流的数字特征和动态概率分布。采用IEEE 14节点和IEEE 118节点系统进行仿真,验证了算法准确性和效率。
Aiming at the conservative property problem of the calculation results of traditional dynamic probability power flow(DPPF),a calculation method of DPPF considering temporal and spatial correlation of forecast error is proposed.Firstly,the correlation coefficient stationary process is used to describe the forecast error process of input variables,and non-parametric kernel density estimation is used to fit the forecast error distribution directly through historical data of forecast error.Afterwards,equivalent possibility transformation and Nataf transformation are used to generate forecast error samples with temporal and spatial correlation.Finally,the Monte Carlo simulation based on Latin hypercube sampling is used for DPPF calculation and the numerical characteristics and dynamic probability distribution of the ode voltage magnitudes and branch power are obtained.The IEEE 14-bus and IEEE118-bus systems are used to verify the precision and efficiency of the proposed algorithm by simulation.
作者
余爽
翁程琳
张程
臧海祥
YU Shuang;WENG Chenglin;ZHANG Cheng;ZANG Haixiang(State Grid Jiangsu Electric Power Co.,Ltd.Maintenance Branch Company,Nanjing,Jiangsu 211100,China;College of Energy and Electrical Engineering,Hohai University,Nanjing,Jiangsu 211100,China;Nanjing University of Aeronautics and Astronautics,Nanjing,Jiangsu 210016,China)
出处
《广东电力》
2021年第10期43-49,共7页
Guangdong Electric Power
基金
国家重点研发计划项目(2018YFB0904500)。
关键词
时空相关性
动态概率潮流
自相关系数平稳过程
非参数核密度估计
拉丁超立方采样
temporal and spatial correlation
dynamic probability power flow(DPPF)
correlation coefficient stationary process
non-parametric kernel density estimation
Latin hypercube sampling
