影响因素重要性筛选条件下的基坑沉降变形预测  

Prediction of Foundation Pit Settlement and Deformation Under Condition of Importance Screening of Influencing Factors

作　　者：彭伟[1] Peng Wei(Shaanxi Railway Institute)

机构地区：[1]陕西铁路工程职业技术学院,陕西渭南714000

出　　处：《勘察科学技术》2022年第4期7-12,共6页Site Investigation Science and Technology

摘　　要：为准确掌握不同影响因素对基坑沉降的重要性程度,实现基坑沉降的高精度预测,先利用层次分析法和模糊理论构建基坑沉降影响因素的重要性筛选模型;再以基坑沉降影响因素的重要性筛选结果为基础,利用优化动态神经网络构建出基坑沉降变形预测模型,以实现沉降变形预测。实例分析表明:在基坑沉降影响因素的重要性筛选过程中,重要性等级属一级的因素有2个,属二级的因素有3个,其余6个因素的重要性等级均为三级;同时,通过沉降变形预测,得出一级影响因素对基坑沉降预测的影响较小,以二级影响因素与三级影响因素组合作为输入层的预测效果相对最优,其预测结果的平均相对误差仅为2.03%,具有较高的预测精度。In order to accurately grasp the importance of different influencing factors on foundation pit settlement and realize the high-precision prediction of foundation pit settlement, the importance screening model of influencing factors on foundation pit settlement is constructed by using analytic hierarchy process and fuzzy theory;Then, based on the screening results of the importance of the factors affecting the foundation pit settlement, the foundation pit settlement deformation prediction model is constructed by using the optimized dynamic neural network to realize the settlement deformation prediction. The example analysis shows that in the process of screening the importance of the factors affecting the settlement of foundation pit, there are 2 factors belonging to the first level of importance, 3 factors belonging to the second level, and the importance of the other 6 factors are all of the third level;At the same time, through the settlement deformation prediction, it is concluded that the primary influencing factors have little influence on the foundation pit settlement prediction, and the prediction effect of taking the combination of secondary influencing factors and tertiary influencing factors as the input layer is relatively optimal. The average relative error of the prediction result is only 2.03%, which has high prediction accuracy.

关 键 词：基坑 沉降变形 层次分析法 重要性筛选 动态神经网络 

分 类 号：TU433[建筑科学—岩土工程]