摘要
文中针对传统GM(1,1)模型背景值求解存在误差,模型预测精度低的问题,通过分析误差来源,根据已有复化Simpson公式优化背景值的方法,利用积分函数分区间积分求解逼近的思想,构建动态序列预测模型,并结合实例分析进一步验证了该方法的有效性,能够减弱背景值影响误差,提高模型的拟合预测精度,对工程实际应用具有适用性。
Aiming at the problem that the traditional GM( 1,1) model had errors in solving the background value and low prediction precision,in this paper,the error sources were analyzed,dynamic sequence prediction model was constructed according to the method of optimizing background value by complex Simpson formula by using the idea of integral function interzonal integration to solve approximation,and the effectiveness of the method was further verified by an example analysis. The influence error of background value could be reduced and the fitting and prediction accuracy of the model was improved,which was applicable to the practical engineering applications.
作者
成枢
周龙飞
朱健
Cheng Shu;Zhou Longfei;Zhu Jian(College of Geomatics,Shandong University of Science and Technology,Qingdao 266590,China)
出处
《矿山测量》
2019年第3期110-113,共4页
Mine Surveying
