摘要
针对化工过程灰箱建模存在的精确度差、速度慢、计算复杂度高等问题,对现行模糊C-均值聚类算法进行了改进,提出了一种快速全局优化的(用于建模的数据训练集)模糊聚类算法.该算法具有不依赖初始条件、收敛速度快等特点.实验结果表明,利用快速全局优化模糊聚类算法得到的数据,在灰色预测的时间和数据准确性方面都有了显著提高,计算机仿真实验表明了该算法的有效性.
The modeling of chemical grey box has the problems of low accuracy,slow and high degree of computational complexity.The Fast Global Fuzzy C-Means Clustering algorithm(FGFC),based on the Fuzzy C-means algorithm,is improved(It is used to get effective training data of modeling).The algorithm does not depend on initial conditions,convergence speed and so on.Experiments show that the FGFCM algorithm's data has improved the performance of predicting time and data accuracy.This algorithm is proved to be efficient by computer simulation.
出处
《哈尔滨理工大学学报》
CAS
北大核心
2010年第3期31-35,共5页
Journal of Harbin University of Science and Technology
基金
黑龙江省科技攻关项目(GZ08A107)
关键词
灰色预测
灰箱建模
模糊聚类算法
全局优化
gray prediction
grey box modeling
Fuzzy C-means clustering
global optimization