摘要
针对灰色模型在建筑物沉降预测中存在的不足,对传统的灰色模型进行改进,提出利用时间差法和Newton插值法将非等间距的原始数据序列转换为等间距数据序列,然后采用GM(1,1)模型和两种加权法对等间距数据进行建模的预测方法。并通过实例进行精度分析,实例证明利用Newton插值法构造等间距数据,采用基于时间加权的模型对沉降进行预测具有较高的精度。采用该模型对建筑物沉降趋势进行预测,为施工决策起到良好的参考作用。
To improve the performance of grey model in predicting buildings settlement,the time difference method and Newton interpolation method were adopted to transform non-equidistant data sequence into equidistant data sequence. The GM（ 1,1） and two weighting methods were employed to predict the buildings settlement. The precision analysis by an example demonstrates that using Newton interpolation method to construct equidistant data combined with the time-weighted model is of high prediction accuracy. The results indicate that the model is applicable in buildings settlement prediction and can contribute to decision-making in construction.
出处
《福建工程学院学报》
CAS
2015年第4期351-356,共6页
Journal of Fujian University of Technology
关键词
预测
灰色模型
沉降
Newton插值法
prediction
gray model
settlement（subsidence）
Newton interpolation method
