土工基础

›› 2015, Vol. 29 ›› Issue (6): 63-66.

• 工程实录 • 上一篇    下一篇

基于BP神经网络的建筑工程地质勘察钻孔质量评价

明前军，姚月亮，张 洁   

  1. (江苏省水文地质海洋地质勘查院，江苏淮安 223003)
  • 收稿日期:2015-06-12 出版日期:2015-12-22 发布日期:2015-12-30
  • 作者简介:明前军（1967-），男，高级工程师，注册岩土工程师，研究方向为岩土工程勘察与设计。

Evaluation of the Borehole Quality in the Geotechnical Investigation by BP Neural Network Process

MING Qianjun, YAO Yueliang, ZHANG Jie   

  1. (Jiangsu Hydrological and Marine Geology Investigation Institute, Huai'an 223003)
  • Received:2015-06-12 Online:2015-12-22 Published:2015-12-30

摘要： 建筑工程地质勘察钻孔质量评价是一项多目标、多因素、多层次的复杂工程。在总结传统影响钻孔质量各类因素的基础上，构建了建筑工程地质勘察钻孔质量综合评价指标体系，综合运用BP神经网络理论建立了钻孔质量BP网络评价模型。该模型能够很好地反映各钻孔质量评价指标与综合评价结果之间复杂的非线性关系，降低钻孔质量评价过程中的主观因素影响。为了验证该评价模型的可靠性与实用性，对评价模型进行了检测并进行实例应用，结果表明可将其用于建筑工程地质勘察钻孔质量评价。

关键词: BP神经网络, 建筑工程, 钻孔质量, 地质勘察

Abstract: The quantitative evaluation of the borehole quality in the geotechnical investigation is a comprehensive system that involves multitargets, multiactors and multilayers. Based on the summary of the boring information evaluation by the conventional methods, a comprehensive borehole quality evaluation system is established by using the Back Propagation (BP) neural Network Process. This system can provide a better nonlinear and complicated relationships between the borehole quality results and the involved factors. This system can also minimize the subjectively biased evaluation results. As an example this BP model is used in a practical borehole evaluation project. The results indicate the application of this model is feasible.

Key words: BP Neural Network, Building Engineering, Borehole Quality, Geotechnical Investigation