基于数列变换优化的灰色系统模型  

Data mapping optimization based grey model

作　　者：付莹[1] 张广立[1] 杨汝清[1] 

机构地区：[1]上海交通大学机器人研究所,上海200030

出　　处：《系统工程与电子技术》2004年第4期430-432,436,共4页Systems Engineering and Electronics

基　　金：国家自然科学基金资助课题(69975014)

摘　　要：提出了一种可对含负项数列建立灰色系统模型的变换方法,并利用Matlab对变换参数进行优化。针对变换函数的特点提出了一种简化变换参数优化的方法,使二维优化问题简化为一维优化,可显著减小优化计算量,同时可获得较高的模型精度,这样就可在处理含负项数列的同时获得较高的模型精度,拓展了GM模型的应用范围。具体算例验证了所提建模方法的有效性。A data mapping method is adopted to get the GM model of data sequence that contains negative items. To get optimum parameters of the data mapping function, optimization is performed through Matlab. Furthermore, based on the characteristic of the data mapping function a simplified data mapping parameters optimization method is proposed for reducing the 2-D to 1-D optimization problem, which can considerably reduce the computational amounts of optimization, and achieve quite high precision of modeling. This method not only can treat data sequence containing negative items, but also can achieve rather precise GM modeling, which enlarge the application areas of GM model. Actual cases validate the efficiency of this method.

关 键 词：灰色系统模型 数列变换 非线性最小二乘 

分 类 号：N94[自然科学总论—系统科学]