数据插值技术对基于遗传编程算法符号回归的影响研究  

The Research of the Influence of Data Interpolating Technology on Symbolic Regression Based on Genetic Programming

作　　者：于鹏[1] 王京 

机构地区：[1]渤海大学工学院,辽宁锦州121000 [2]国网锦州供电公司,辽宁锦州121000

出　　处：《江西师范大学学报（自然科学版）》2014年第3期286-289,共4页Journal of Jiangxi Normal University(Natural Science Edition)

基　　金：辽宁省科学事业公益研究基金(2011004001)资助项目

摘　　要：提出一种先采用样条函数插值对数据进行平滑处理,再采用基于遗传编程的符号回归方法进行数据分析的新方法.与实验补充数据的方法相比,该方法实验成本低、风险小、受环境影响小,在电力电子技术领域具有广泛的适用性和较好的应用潜力.A new method which use data interpolation technology to preprocessing data has been proposed. After da-ta preprocessing,the data is processed by the symbol regression method. Comparing to traditional method that get data by practical experiment,this method is cheap and low-risk. The method is independent from environment. In the field of power electronics,this method has strong applicability and a good application prospect.

关 键 词：符号回归 样条插值 遗传编程 数据拟合 电力电子 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]