摘要
针对隶属度函数及其确定方法研究中存在的问题,通过采用高斯函数和sigmoid函数表示隶属度函数,提出了基于密度指标的单次聚类方法,它的计算量仅与输入数据数目成线性关系,且能快速得到聚类中心和数据的隶属度;在实例仿真中,确定了鄂北石油钻井现场数据EB_Da51的隶属度函数,解决了隶属度函数难以确定的难题.
The research of membership function and its determination method is of great significance.For determining the membership function expressed by Gaussian or sigmoid function,this paper develops a density index based on single-shot clustering method.The calculation amount depends on the number of input data linearly and the center and the membership degree of engineering data can be easily obtained.Through simulation,membership function of the data EB_Da51 from the petroleum drilling in Ebei is established,and the problem that membership function is hard to be described has been solved in paper.
出处
《郑州大学学报(工学版)》
CAS
北大核心
2012年第2期28-30,共3页
Journal of Zhengzhou University(Engineering Science)
基金
国家自然科学基金资助项目(60974005)
河南省自然科学研究计划项目(2010A120008)
关键词
隶属度函数
单次聚类
曲线拟合
membership function
single-shot clustering
curve fitting