基于神经网络的桩基完整性预测分析

反射波法进行桩基完整性检测时，常规理论分析结果与实际状态存在一定的差异。文中以反射波法检测基本数据为基础，分析影响桩基质量的主要因素，针对桩土共同作用的动静力学特征，选择了5个方面的重要因素，建立神经网络评价系统的影响因子，根据这些影响因素对桩基质量检验的影响规律，给出它们的影响系数。通过建立神经网络预测系统，对桩身缺陷的位置、类型和程度给予合理准确的解释。

Prediction Analysis on Pile Foundation Integrity based on Neural Network

There is a certain difference between the results of conventional theory analysis and actual state in pile foundation integrity detection with reflected wave method.This paper analyzes the main factors that influence pile foundation quality based on the basic detected data attained by reflected wave method and chooses 5 major factors to establish influence factors of neural network evaluation system in accordance with dynamic and static mechanic characteristics of pile soil interaction;then gives influence coefficient according to the regularity of influence from these factors on pile foundation quality detection.Through establishing neural network prediction system,it gives reasonable and exact explanation of the position,type and degree of pile body defect.