灰色时序模型在建筑物沉降观测中的应用  

Application of Grey Time-series Model in Building Settlement Observation

作　　者：潘顺宇 付饶 PAN Shunyu;FU Rao(Shandong Zhengyuan Digital City Construction Co.,Ltd.,Yantai 264670,China)

机构地区：[1]山东正元数字城市建设有限公司,山东烟台264670

出　　处：《地理空间信息》2022年第10期147-150,154,共5页Geospatial Information

摘　　要：灰色时序模型是建筑沉降预测的常用模型,具有原始序列需求少、预测时间长的优点,但在原始序列的长期预测中通常存在精度偏低的问题。为验证梯形公式在背景值优化中的作用,进一步使用牛顿柯达斯系数下的其他数值积分公式进行背景值优化,并针对灰色模型的残差使用傅里叶级数进行残差修正。实例计算表明最佳模型为傅里叶级数残差修正下n=3的数值积分改进模型,平均误差为1.82 mm,符合精度要求。Grey time-series model is a common model for building settlement prediction,which has the advantages of less demand for original se-ries and long prediction time,but it usually has the problem of low accuracy in the long-term prediction of original series.In this paper,in order to verify the action of trapezoidal formula in the background value optimization,we further used other numerical integration formulas under the Newton Kodak coefficient to optimize the background value,and used Fourier series to correct the residual of gray model.The example calcula-tion result shows that the best model is the improved numerical integration model of n=3 under the residual error correction of Fourier series,and the average error of offset is 1.82 mm.

关 键 词：灰色模型 数值积分 傅里叶级数 背景值优化 沉降预测 

分 类 号：P258[天文地球—测绘科学与技术]