基于精英策略的气象数据协同进化属性约简  

Attribute Reduction Algorithm Based on Elite Strategy for Co-Evolution of Meteorological Data

作　　者：程勇 郑忠仁[2] 王军 CHENG Yong;ZHENG Zhong-ren;WANG Jun(Technology Industry Department,Nanjing University of Information Science & Technology,Nanjing 2 Jiangsu 210044,China;Department of Computer & Software,Nanjing University of Information Science & Technology,Nanjing Jiangsu 210044,China)

机构地区：[1]南京信息工程大学科技产业处,江苏南京210044 [2]南京信息工程大学计算机与软件学院,江苏南京210044

出　　处：《计算机仿真》2019年第8期231-235,252,共6页Computer Simulation

基　　金：国家自然科学基金资助项目(61402236,61373064);赛尔网络下一代互联网技术创新项目(NGII20160318);江苏省“六大人才高峰”项目(2015-DZXX-015)

摘　　要：针对气象数据属性冗余度高和现有属性约简算法效率慢的问题,提出一种基于精英策略的协同进化属性约简算法。该算法将进化种群分为两个子种群,一个子种群借助精英个体协助交叉,提高算法的收敛速度。另一子种群通过引入随机种群,平衡进化过程中种群多样性,最后两个子种群协同完成进化操作。与TSDPSO-AR算法和ARAGA算法对气象数据进行降水属性约简操作,结果表明,提出的算法维持了进化过程中种群的多样性,提高了约简性能,简化了信息系统。Aiming at the problem of high attribute redundancy for meteorological data and slow efficiency of existing attribute reduction algorithms, an attribute reduction algorithm based on elite strategy for co-evolution of meteorological data was proposed. By dividing the evolutionary population into two subpopulations, a subpopulation used the elite individuals to assist crossover operations to increase the convergence speed of the algorithm, another subpopulation balanced population diversity through evolution by introducing random populations, and the last two subpopulations completed the evolutionary operations cooperatively. With the TSDPSO-AR algorithm and ARAGA algorithm , the precipitation attribute reduction operation for meteorological data was performed. The results show that the proposed algorithm maintains the diversity of population during evolution, improves the reduction performance, and simplifies the information system.

关 键 词：气象数据 属性约简 精英策略 协同进化 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] TP301.6[自动化与计算机技术—控制科学与工程]