兰州市水源地建设工程初始地应力场反演方法研究

发布时间：2018-05-09 02:05

本文选题：岩体工程 + 输水隧洞　；参考：《东华理工大学》2017年硕士论文

【摘要】：随着工程建设的发展,越来越多的重大工程在深部和强烈构造活动区兴建,给工程的安全建设提出了重大挑战。由于初始地应力是分析围岩的稳定性并实现工程开挖设计和科学决策的重要条件。因此,准确测量和反演初始地应力是进行工程灾害风险评估、灾害预警以及可靠防控的前提和关键。针对目前地应力反演方法中的反演精度问题,本文以兰州水源地建设工程为工程背景,基于工程现场初始地应力实测资料,提出了一种结合多元线性回归方法和PSO-BP神经网络方法的岩体初始地应力场反演的新方法。本文的主要研究工作如下:(1)对初始地应力的研究背景进行了介绍,并归纳总结了国内外地应力测量方法和技术的适用性及发展趋势;阐述了目前各种地应力场反演分析方法及各自的优缺点,同时并介绍了初始地应力场反演的基本原理。(2)介绍了兰州水源地工程区域的工程概况与地质条件,并基于现场地应力的实测数据,使用多元线性回归、非线性回归等反分析方法对兰州水源地工程区域的初始地应力场进行反演分析。(3)提出了基于线性回归的PSO-BP神经网络反演新方法,即基于多元线性回归地应力反演结果重新构造样本数据,以此开展基于PSO-BP神经网络模型的地应力反演,使用该方法对兰州市水源地建设工程区域的初始地应力场进行反演分析。同时,为验证新方法的合理性,采用工程经验类比法构造了经验样本数据,并进行了初始地应力的神经网络反演。(4)将上述四种地应力场反演方法得到的工程区域初始地应力计算值分别与实测值对比分析可得到:相对其他方法,本文所提新方法不论在地应力量值还是方向性方面均能与实测地应力数据很好吻合,尤其在地应力方向性方面具备更好的准确性。本文提出的基于线性回归方法的PSO-BP神经网络新方法在初始地应力场反演中能够考虑多种地质构造作用的荷载或位移作用模式和自重的影响,避免现有方法中构造样本时因为经验不足而导致的错误,此外,通过回归与回代过程,避免了试算法受人为因素影响和随意性较大的缺点并有效地保证工程区域初始地应力场的准确反演。本文研究成果可以为相似工程的初始地应力场反演提供重要参考。
[Abstract]:With the development of engineering construction, more and more important projects are built in deep and strong tectonic active areas, which brings great challenges to the safety construction of engineering. The initial in-situ stress is an important condition for analyzing the stability of surrounding rock and realizing engineering excavation design and scientific decision. Therefore, accurate measurement and inversion of initial in-situ stress is the prerequisite and key for engineering disaster risk assessment, disaster early warning and reliable prevention and control. Aiming at the problem of inversion accuracy in the present inversion method of in-situ stress, this paper takes Lanzhou water source construction project as the engineering background, and based on the measured data of the initial in-situ stress in the engineering site. A new method for inversion of initial geostress field of rock mass by combining multiple linear regression method and PSO-BP neural network method is proposed. The main research work of this paper is as follows: (1) the research background of initial in-situ stress is introduced, and the applicability and development trend of in-situ stress measurement methods and techniques at home and abroad are summarized. This paper expounds various methods of inversion and analysis of in-situ stress field and their advantages and disadvantages, and introduces the basic principle of inversion of initial in-situ stress field.) it also introduces the engineering situation and geological conditions of Lanzhou water source engineering area. Based on the measured data of in-situ stress, multivariate linear regression is used. A new inversion method of PSO-BP neural network based on linear regression is proposed for inversion analysis of initial in-situ stress field in Lanzhou water source engineering area by nonlinear regression and other inverse analysis methods. In other words, the sample data are reconstructed based on the results of multiple linear regression geostress inversion, and the in-situ stress inversion based on PSO-BP neural network model is carried out, and the initial in-situ stress field of the water source construction project in Lanzhou City is analyzed by using this method. At the same time, in order to verify the rationality of the new method, the empirical sample data are constructed by using the engineering empirical analogy method. The neural network inversion of initial in-situ stress is also carried out. (4) the calculated values of the initial in-situ stress obtained by the four methods mentioned above are compared with the measured values, and the results are as follows: compared with other methods, The new method proposed in this paper is in good agreement with the measured in-situ stress data in terms of both the in-situ stress value and the directionality, especially in the directionality of the in-situ stress. In this paper, a new PSO-BP neural network method based on linear regression method is proposed, which can take into account the influence of load or displacement modes and deadweight of various geological structures in the inversion of initial geostress field. To avoid errors caused by inexperience in constructing samples in existing methods, in addition, through regression and retrogeneration processes, The disadvantages of artificial factors and randomness of the trial algorithm are avoided and the accurate inversion of the initial geostress field in the engineering area is effectively ensured. The results of this paper can provide an important reference for the inversion of initial geostress field of similar engineering.
【学位授予单位】：东华理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TV223

【参考文献】