房屋建筑桩基工程施工质量检测技术研究

Research on Construction Quality Detection Technology of Building Pile Foundation Engineering

研究一种基于神经网络数据深度迭代回归方法的房屋建筑桩基础施工质量检测方法。根据桩基础施工过程中的大孔径深孔钻机和套管沉管机的实际运行数据,对其进行基于神经网络对数回归函数模块和二值化重投影回归函数模块的深度迭代回归分析,得到精度在10 mm的桩基础解析度评价结果。此数据可以较传统微震测试法给出更加清晰的异常区边界且给出异常区内部数据细节,但新方法仍缺少对桩基础可用性的决策性评价方法,需要与传统检测方法联合使用。

A construction quality detection method of building pile foundation based on neural network data depth iterative regression method has been studied.The actual operation data of large diameter deep hole drilling rig and casing pipe sinking machine in the process of pile foundation construction is analyzed.In addition,the depth Iterative regression analysis of logarithmic regression function module and binary reprojection regression function module is carried out.Thus the evaluation result of pile foundation resolution with accuracy of 10 mm is obtained.Compared with the traditional microseismic test method,the boundary of the abnormal area is more clear,and the internal data of the abnormal area is more detailed.However,the new method still lacks the decision-making evaluation method for the availability of pile foundation,which needs to be used in combination with the traditional detection method.