基于K-means聚类算法的城市轨道交通车站配电变压器容量计算新方法

目前城市轨道交通车站配电变压器容量的传统计算方法是需要系数法,但此方法的计算结果与实际运行需求相差较大。应用K-means聚类算法,通过对已运营车站历史数据进行分析,形成标准负荷曲线,结合新建车站设备的设计功率,最终确定新建车站的变压器容量。应用此新方法,能显著提高变压器计算结果与实际运行的负载率匹配性,大幅降低配电变压器的计算容量,从而明显减少配电系统一次性投资和后期运营成本。

New Method for Urban Rail Transit Station Distribution Transformer Capacity Calculation Based on K-means Clustering Algorithm

At present, the traditional calculation method of urban rail transit station distribution transformer capacity is a demand coefficient method. However,the results of this method fail to meet the actual operational requirements. In this paper,the K-means clustering algorithm is used based on the analysis of historical data of operating stations and a standard load curve is formed with reference to the new designed power of station equipment to determine the capacity of the new transformer station. Using this new method significantly improves the transformer load matching rate between calculation results and actual operation load and greatly reduces the calculation capacity of distribution transformers,thereby significantly reduces the one-time investment and operating costs of distribution system.