【摘要】 为对配电网的发展状况进行客观、高效的综合评价,提出一种基于非线性映射与聚类分析的配电网评价方法。首先运用遗传算法与非线性映射法将高维配电网评价指标降维成二维数据;然后对二维数据进行K-means聚类分析,并用模拟退火算法优化聚类结果;最后结合用因子分析法构建的配电网评价指标体系,对聚类结果进行分析,从经济性、安全性、可靠性3个角度评价配电网的发展状况。将本文提出的方法应用于河南省县域配电网的综合评价,客观地反映了其发展现状及主要问题,表明本文所用方法是科学有效的。更多还原

【Abstract】 To objectively and efficiently evaluate the development of distribution network,an evaluation method for distribution network is proposed based on nonlinear mapping and clustering analysis. First,genetic algorithm and non-lin-ear mapping method are used to reduce the high-dimensional evaluation indexes of distribution network to 2-dimension-al data. Then,K-means clustering analysis is carried out on the 2-dimensional data,and the clustering results are opti-mized using the simulated annealing algorithm. Finally,the clustering result is analyzed with the combination of a distri-bution network evaluation index system constructed using the factor analysis method,which can evaluate the develop-ment status of distribution network from three aspects,i.e.,economic efficiency,security,and reliability. The methodproposed in this paper is applied to the comprehensive evaluation of county-level distribution networks in Henan Prov-ince,and the clustering results reflect the development status of the distribution networks together with their main problems,showing that the proposed method is scientific and effective.更多还原