基于AP聚类的含分布式光伏的配电网分区方法

Distribution network partition method based on AP clustering with distributed photovoltaic

分布式光伏(Distributed Generation Photovoltaic,DGPV)接入配电网(Distribution Network,DN),其出力具有随机性和波动性,为配电网分区带来了挑战。为获得稳定的全网分区,提出了一种基于AP聚类的分布式光伏接入配电网的分区方法。首先,针对分布式光伏接入配网其出力对电压的影响,提出了一种适用于除平衡节点外所有节点的综合电气距离指标;其次,考虑以离散概率分布处理光伏出力,建立节点间电气距离期望矩阵作为分区依据,采用AP聚类对除平衡节点外的所有节点进行分区,以IEEE-123节点系统对上述方法进行了验证,结果表明所提方法的有效性和可行性。该分区方法优于以往方法,降低了将PQ节点与PV节点分别进行分区处理的复杂度,且避免了人为确定聚类数目带来的人为主观性,并将分布式光伏出力的随机性和波动性考虑在内,具有广泛的适用性。

The distributed distribution photovoltaic(DGPV)access distribution network(DN),its output has randomness and volatility,which brings challenges to the distribution network partition.In order to obtain a stable full-network partition,this paper proposes a partition method based on AP clustering for distributed PV access distribution network.Firstly,for the distributed photovoltaic access distribution network,the effect of its output on the voltage,an integrated electrical distance indicator applicable to all nodes except balanced nodes is proposed.Secondly,considering the distribution of photovoltaic power by discrete probability distribution and establishing the electrical distance expectation matrix between nodes as the partition basis,AP clustering is used to partition all nodes except balanced nodes,and the above method is verified by IEEE123-node system.The results show the effectiveness and feasibility of the proposed method.The partition method is superior to the previous method,reducing the complexity of partitioning the PQ node and the PV node respectively,and avoiding the human subjectivity caused by artificially determining the number of clusters,and taking the randomness and fluctuation of the distributed photovoltaic output into account,which have wide applicability.