基于聚类算法的电网工程数据分析方法

Data analysis method of power grid engineering based on clustering algorithm

针对数据分析技术在电网工程数据测算分析领域的应用问题,文中分析了电网工程的技术特征,采用层次分析法(AHP)构建了电网工程的技术参数体系,并进一步提出了基于层次凝聚聚类(HAC)与多元回归分析(MRA)的电网工程数据分析方法。文中所提的方法是采用HAC方法对电网工程数据进行聚类,采用MRA方法对聚类后的数据进行分析,得到回归方程用于工程造价数据的测算,使得特性相近的数据划分为同组,避免特征差异较大的数据影响测算准确性。通过110 kV变电工程与架空线路输电工程的验算表明,所提方法的工程造价数据测算误差均在10%以内,能够准确地进行电网工程造价测算。

Aiming at the application of data analysis technology in the field of power grid engineering cost calculation.This paper analyzes the technical characteristics of power grid engineering,constructs the technical parameter system of power grid engineering by Analytic Hierarchy Process(AHP),and further proposes the data analysis method of power grid engineering based on Hierarchical Agglomerative Clustering(HAC)and Multiple Regression Analysis(MRA).The method proposed in this paper uses the HAC method to cluster the power grid engineering data,and the MRA method to analyze the clustered data and obtain the regression equation for the cost calculation,so that the data with similar characteristics are divided into the same group to avoid the data with large characteristics differences affecting the accuracy of cost calculation.Through the calculations of 110 kV substation project and overhead line transmission project,the results show that the cost estimation error of the proposed method is within 10%,which can accurately calculate the cost of power grid engineering.