聚类算法在多变压器选址中的应用研究

Research on clustering algorithm in transformer location

为有效解决一定区域内多台变压器选址问题,提出了一种聚类算法下基于导线选型策略的多变压器选址优化模型。该模型根据电力负荷的分布情况和变压器数量,确定变压器的初始位置,再利用负荷点与变压器位置的远近进行聚类调整。通过对变压器的位置优化和供电范围进行合理分配,可以进一步降低线损及提高经济效益。在运用聚类算法的同时,还引入了实际情况下的电缆的选型策略,使它接近实际情况。将该算法应用到一个算例中,其计算结果与传统方法的比较结果表明,所提模型在多变压器优化选址和降低变压器线损方面均优于传统方法。

To solve the transformer location problem in an area effectively,an optimization model based on conductor selection strategy of clustering algorithmis proposed.Thismodel determined the initial location of transformers by the power load and the number of transformer first,and then adjusted the location by using the distances of the transformer and the load point.By optimizing the location and power supply area of transformer,the loss is reduced.Not only the clustering algorithm is used,but also the type selection strategy of cables is adopted for purpose of close to the actual situation.Asimulation example is used to verify this algorithm,which is compared to traditionalmethod,indicated that this algorithm is better than traditional method both in optimal location of transformers and reducing line loss.