摘要
针对传统的分段曲线拟合方法在选择拟合函数和确定分段区间时经验成分较多的不足,提出一种自动分段多项式曲线拟合方法,根据误差方差和误差均值,自动确定经验函数和分段区间。通过实际数据的检验,验证了该方法的拟合效果。
Aiming at the deficiency of traditional piecewise curve-fitting method which contains many experiential elements in the process of selecting fitting function and separating data into pieces,this paper puts forward a automatically piecewise polynomial curve-fitting method which can finds out empirical function and piecewise interval automatically according to the error variance and error mean. Through a group of actual data,the effect of this method is validated.
出处
《科学技术与工程》
北大核心
2014年第3期55-58,共4页
Science Technology and Engineering
基金
华为公司创新研究计划(YB2012120202)资助
关键词
数据拟合
分段拟合
多项式曲线
最小二乘法
data-fitting piecewise-fitting polynomial curve least-square principle
