基于模糊聚类分析的能源产业信息自动挖掘建模研究  

Research on automatic mining and modeling of energy industry information based on fuzzy cluster analysis

作　　者：陈竞 杜杰 丁胜利 CHEN Jing;DU Jie;DING Shengli(Southern Power Grid Digital Grid Research Institute Co.,Ltd.,Guangzhou 510663,China)

机构地区：[1]南方电网数字电网研究院有限公司,广东广州510663

出　　处：《电子设计工程》2024年第1期138-141,共4页Electronic Design Engineering

摘　　要：针对当前能源产业信息挖掘结果完整性差的问题,提出一种基于模糊聚类分析的能源产业信息自动挖掘建模方法。在模糊聚类分析算法中引入核学习算法,添加Gaussian核函数,搭建核模糊分析算法。确定核模糊分析算法的初始聚类中心,利用粒子群优化算法优化初始聚类中心,根据优化的初始聚类中心建立能源产业信息自动挖掘的目标函数,获取适应度值。根据适应度值与目标函数选择最佳个体,求解最佳个体的聚类有效性函数,解码输出聚类有效性函数最大时所对应的最优聚类数量与对应的聚类中心,以此搭建能源产业信息自动挖掘模型。实验结果表明,该模型可有效挖掘能源产业信息,在数据集规模不同的情况下该模型的调整兰德系数均较高,挖掘结果的完整性较高,自动挖掘效果佳。Aiming at the problem of poor integrity of current energy industry information mining results,an automatic mining and modeling of energy industry information based on fuzzy cluster analysis is proposed.The kernel learning algorithm is introduced into the fuzzy clustering analysis algorithm,and the Gaussian kernel function is added to build the kernel fuzzy analysis algorithm.Determine the initial cluster center of kernel fuzzy analysis algorithm,optimize the initial cluster center by particle swarm optimization algorithm,and establish the objective function of energy industry information automatic mining according to the optimized initial cluster center to obtain the fitness value.According to the fitness value and the objective function,the best individual is selected,the clustering effectiveness function of the best individual is solved,and the optimal cluster number and the corresponding cluster center corresponding to the maximum clustering effectiveness function are decoded,so as to build an energy industry information automatic mining model.The experimental results show that the model can effectively mine energy industry information.The Adjusted Rand Index of the model is high in different data sets,the integrity of the mining results is high,and the automatic mining effect is good.

关 键 词：模糊聚类分析 能源产业信息 自动挖掘 密度聚类 粒子群 有效性函数 

分 类 号：TN-9[电子电信]