摘要
工程中对得到的实验数据进行拟合,常采用最小二乘法,拟合函数多采用代数多项式。但当多项式次数比较大时,会出现病态问题,从而导致计算结果误差增大,于是在拟合时应避免产生过高次数的多项式,就有必要确定次数较低的最佳拟合。针对基本初等函数,论述确定最佳拟合的方法,并结合VC++和Matcom混合编程,快速准确实现最佳拟合。通过分析实验结果,证明了方法的有效性。
When fitting to the test data in projects, least square method is often adopted and the algebra polynomial is often used as the fitting function. But there is a morbid state when the polynomial counts become bigger, and then the result error become bigger, so de- termining the best fitting is necessary. The method of determination of the optimal power of data fitting is introduced and a program with hybird programming ofVC++ and Matcom is realized. Through analyzing the experimental result, themethodprovesefficientandvaluable.
出处
《计算机工程与设计》
CSCD
北大核心
2007年第12期2980-2982,共3页
Computer Engineering and Design
基金
北京市高等学校教育教学改革立项基金项目(2005075)
北京市教委科技发展计划面上基金项目(KM200610028013)
北京市自然科学基金项目(4062009)
关键词
数据拟合
最小二乘法
代数多项式
最佳拟合
混合编程
data fitting
least square method
algebra polynomial
optimal fitting
hybrid programming