基于遗传规划的深基坑围护结构侧移预测方法

发布时间：2018-04-24 10:32

本文选题：深基坑 + 围护结构　；参考：《华中科技大学》2015年硕士论文

【摘要】：我国目前基础建设工程快速发展,大量工程建设都与基坑密切相关。基坑工程具有复杂性和不可预计性的特点,研究基坑围护结构的侧移预测能为基坑的安全性和稳定性提供参考,促进对基坑的设计和施工进行优化,有利于预防和减少基坑相关事故,降低人员伤亡和财产损失。本文研究分析了目前国内外基坑围护结构变形的智能算法,比较并总结这些算法运用在基坑工程中的优势和局限性;引出遗传规划理论并分析其原理和特点,在此基础上提出运用遗传规划模型预测基坑围护结构的变形。本文还介绍了基坑围护结构变形机理和类型,着重探讨开挖深度、开挖阶段、平面尺寸、支撑刚度、围护长度和厚度等影响围护结构侧向位移的因素。重点研究遗传规划模型的建立和各参数设置,主要包括适应度函数、选择概率、交叉概率、变异概率等。同时利用工程数据对模型进行训练和测试,筛选出最终结果表达式,与神经网络模型所得结果进行对比,分析两者差异性,检验本表达式的精度要求。最后利用武汉地铁深基坑实际开挖数据对最终结果进一步验证,证明本模型在基坑围护结构侧移预测方面的可靠性和实用性。本文的研究涉及基坑工程领域和计算机编程领域,使用新的角度将不同的领域有机结合起来,为深基坑围护结构的侧向变形预测提供了新的解决思路,推动了遗传规划在基坑领域的实际工程运用,并具有较高运用价值。同时对于基坑安全和创造经济效益更多价值具有重要的现实意义,能够为基坑工程变形和为类似工程提供有益的参考。
[Abstract]:With the rapid development of infrastructure engineering in China, a large number of projects are closely related to foundation pit. Foundation pit engineering has the characteristics of complexity and unpredictability. The study of lateral displacement prediction of foundation pit enclosure structure can provide a reference for the safety and stability of foundation pit and promote the optimization of the design and construction of foundation pit. It is helpful to prevent and reduce the accidents related to foundation pit and reduce casualties and property losses. This paper studies and analyzes the intelligent algorithms for deformation of foundation pit envelope structure at home and abroad, compares and summarizes the advantages and limitations of these algorithms in foundation pit engineering, leads out the theory of genetic programming and analyzes its principle and characteristics. On the basis of this, genetic programming model is used to predict the deformation of foundation pit retaining structure. This paper also introduces the deformation mechanism and types of foundation pit enclosure structure, and emphatically discusses the factors that affect the lateral displacement of the retaining structure, such as excavation depth, excavation stage, plane size, support stiffness, length and thickness of enclosure. This paper focuses on the establishment of genetic programming model and the setting of parameters, including fitness function, selection probability, crossover probability, variation probability and so on. At the same time, the model is trained and tested by engineering data, and the final result expression is screened out, and compared with the result obtained by neural network model, the difference between them is analyzed, and the accuracy requirement of this expression is tested. Finally, the final results are further verified by the actual excavation data of the deep foundation pit of Wuhan Metro, which proves the reliability and practicability of the model in predicting the lateral displacement of the retaining structure of the foundation pit. The research in this paper involves the field of foundation pit engineering and computer programming, and organically combines different fields with new angles, which provides a new solution for lateral deformation prediction of deep foundation pit retaining structure. It promotes the practical engineering application of genetic programming in the field of foundation pit and has higher application value. At the same time, it has important practical significance for foundation pit safety and creating more value of economic benefit, and can provide beneficial reference for foundation pit engineering deformation and similar engineering.
【学位授予单位】：华中科技大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TU753

【相似文献】