基于自适应扩散核密度估计的时序相关概率最优潮流计算方法

为准确评估光伏与负荷的时序性和相关性对电力系统运行状态的影响,提出一种基于自适应扩散核密度估计的时序相关概率最优潮流计算方法。首先,利用光伏出力的自适应扩散核密度估计模型将高斯核函数转换为线性扩散过程,采用渐进积分误差法(asymptotic mean integrated squared error,AMISE)为扩散核函数选取自适应最优带宽,提高了光伏出力模型的局部适应性;其次,利用Copula理论构建光伏与负荷的时序联合概率分布模型,并获取具有相关性的时序光伏出力与负荷样本,进而提出能够准确计及光伏与负荷时序性和相关性的概率最优潮流计算方法;最后基于我国某地光伏电站实测数据与IEEE30节点系统进行仿真分析,验证了所提出计及光伏出力与负荷时序相关性的概率最优潮流计算方法的准确性与有效性。

Probabilistic Optimal Power Flow Considering Correlation and Time Series Based on Adaptive Diffusion Kernel Density Estimation

In order to accurately evaluate the influence of the sequential feature and correlation of photovoltaic output(PV) and load on the operation state of power system, a time-series correlation probability optimal power flow calculation method based on adaptive diffusion kernel density estimation was proposed. Firstly, using the adaptive diffusion kernel density estimation model of photovoltaic output, the Gaussian kernel function was transformed into a linear diffusion process, and the Gaussian kernel function was transformed into a linear diffusion process. In order to improve the local adaptability of the PV output model, the adaptive optimal bandwidth was selected for the diffusion kernel function by using the asymptotic mean integrated squared error(AMISE) method. Secondly, Copula theory was used to build the joint probability distribution model of time series of PV output and load, and relevant samples of time series PV output and load were obtained. Genetic algorithm was used to calculate the probabilistic optimal power flow which could accurately account for the time series and correlation of photovoltaic and load. Finally, based on the measured data of a PV power station in China and the simulation analysis of IEEE30 system, the accuracy and validity of the proposed probabilistic optimal power flow calculation method considering the correlation between PV output and load sequence are verified.