非等时距GM(1,1)在变形预报中的应用  

Application of non-equal interval GM(1,1) in deformation forecast

作　　者：佘娣[1] 谢劭峰[1] 彭家頔[1] 刘燕芳[1] 

机构地区：[1]桂林理工大学测绘地理信息学院/广西空间信息与测绘重点实验室,广西桂林541004

出　　处：《测绘科学》2013年第3期110-111,86,共3页Science of Surveying and Mapping

基　　金：国家自然科学基金项目(41071294;41064001);广西科学基金资助项目(桂科基0991023;0728097)

摘　　要：本文应用灰色理论建模原理,探讨了非等间距GM(1,1)模型的建模步骤;通过几种不同参数下的建模方式,利用MATLAB编程进行计算比较,着重分析影响非等间距GM(1,1)模型预报精度的因素,并探讨模型的改进方法。结果表明,对于较均匀的非等间距数据,灰色模型具有较高的预报精度。在间距间隔变化较大的情况下,灰色模型进行改进后也能得到比较满意的预报结果。On the basis of Gray theoretical modeling principles, the paper discussed the modehng steps oi non-equal interval GM （1, 1 ）. Through several modeling methods with different parameters, the calculation and comparison were conducted using MATLAB programming. It focused on analyzing the impact factors on the accuracy of non-equal GM （ 1, 1 ） model forecast, and explored ways to improve the model. The results showed that, for relatively uniform unequally spaced data, the gray model has higher forecast precision. In the case of larger spacing interval changes, improved gray model could also get satisfactory forecast results.

关 键 词：非等间距 灰色模型 新陈代谢 预报 

分 类 号：P258[天文地球—测绘科学与技术]