基于长短期记忆网络的电网概率潮流计算及态势感知

Probabilistic Power Flow Calculation and Situation Awareness Based on Long Short Term Memory Network

可再生能源发电大量地并入电力系统发输配用各个部分,使配电网呈现主动性,由于电力系统发电的随机性、波动性特点,使潮流大小无法准确预估。对此,提出了基于长短期记忆网络的电网概率潮流预估及态势感知方法。在前人研究的基础上,提出了风力发电、光伏发电、需求侧负荷、电动汽车充电、发电机组的功率概率模型;进一步,基于Nataf方法将多种概率模型进行去相关标准正态分布变换,实现发电负荷的统一;然后,建立了多方协调互补、成本最小的概率调度模型,并使用长短期记忆网络进行概率潮流求解;最后,以某实际电网为例,对所提算法进行验证比较,证明了所提方法的有效性。

A large number of renewable energy power generation are incorporated into various parts of power system transmission and distribution,which makes the distribution network active.Due to the randomness and volatility of power generation in the power system,the power flow cannot be accurately estimated.Therefore,a probabilistic power flow prediction and situation awareness method based on long short-term memory network is proposed.On the basis of previous studies,the power probability models of wind power generation,photovoltaic power generation,demand side load,electric vehicle charging and generator set are proposed.Furthermore,based on Nataf method,decorrelation standard normal distribution transformation is carried out for various probability models to realize the unification of generation load.Then,a probabilistic scheduling model with multi-party coordination and complementarity and minimum cost is established,and the probabilistic power flow is solved by using long short-term memory network.Finally,an actual power grid is taken as an example to verify and compare the proposed algorithm,which shows the effectiveness of the proposed method.