【作者】 撒文奇；

【导师】 张社荣；

【作者基本信息】 天津大学 ， 水工结构工程， 2013， 博士

【摘要】 地下洞室群施工是一项复杂的系统工程，会受到各种复杂天然地质状况等诸多未知因素的影响。针对地下洞室群施工过程的随机性、突发性、不确定性和经验性强等特点，本文引入物联网等先进的技术与方法，从信息采集、工程进度、工程安全、工程预警四个主要方面对地下洞室群施工过程开展了系统地分析研究，研发基于物联网技术的地下洞室群施工专家信息系统。本文主要研究内容和成果如下：（1）基于物联网技术，给出了地下洞室群各种施工信息基于物联网的实时采集建设方案，搭建了地下洞室群施工信息实时采集平台。本文首次将物联网技术应用到地下洞室群施工信息采集中，结合地下洞室群施工过程中产生各类信息的各自特点，建立三种不同的信息采集方式，实现地下洞室群施工过程中信息的无线实时采集，弥补了目前地下洞室群施工信息采集中信息采集过程复杂、时效性差等不足之处，为地下洞室群施工过程中科学管理与决策提供即时的数据信息保障。（2）提出了基于贝叶斯理论的地下洞室群时变施工进度风险预测方法，同时结合施工进度风险控制理念，研发了基于贝叶斯理论的地下洞室群施工进度风险实时控制专家系统。本文将贝叶斯理论与实时仿真技术同时引入到地下洞室群施工进度预测中，建立了基于贝叶斯理论的地下洞室群时变施工进度风险预测方法。该方法将工序单位工作量施工时间的均值和方差视为随机变量，利用现场采集到的实际施工记录信息，不断修正和改进正在施工工序剩余工作时间和未施工工序工作时间的预测概率分布，同时结合柔性网络计划仿真方法实现对工程后期整体施工进度的完工风险预测。基于本文方法得到的施工进度预测结果更加贴近工程实际，对于地下洞室群施工进度概率预测方法的发展起到了一定的推动作用。（3）给出了适用于实时仿真的数值模型建立方法，建立实际施工状态与数值模型模拟状态的实时映射关系，提出了基于数值模拟技术的地下洞室群施工结构安全实时数值仿真方法。本文通过对地下洞室群实际施工状态与施工数值模拟状态的深入分析，给出各自所包含的主要要素，并引入数学中集合的概念，建立了两个状态之间的实时动态映射关系，完成了实际施工进度、突发不良地质及支护措施在大尺度精细化数值模型上的实时、自动化映射模拟，实现了地下洞室群施工过程中结构安全的实时数值仿真，为现场工程师进行指挥决策提供了重要的参考依据。（4）将（2）、（3）两个创新点进行融合，提出了地下洞室群施工期结构安全与进度耦合实时仿真方法，研发了基于结构安全与进度耦合实时仿真的地下洞室群施工突发不良地质处理措施专家系统。大型地下洞室群施工安全与进度受突发不良地质状况影响较大，实时确定不良地质对施工结构安全与进度的综合影响具有重要工程意义。本文将突发不良地质状况映射到数值模型上，根据实时数值仿真结果对不良地质段施工的结构安全性进行预测评估，对不满足安全要求的施工状态提出相应的处理措施，并对处理效果进行实时仿真预测，将确定的处理措施以施工参数的形式反馈给施工进度控制系统，实时评价处理措施对总体施工进度的影响。该系统实现了地下工程安全施工与进度控制的现场办公一体化，为地下工程现场施工的安全与进度控制提供了技术支持。（5）基于强度折减的思想，结合（3）中的实时数值仿真方法，提出了基于实时数值仿真的地下洞室群施工期围岩劣化损伤折减计算方法，该方法能够给出施工期某一施工状态下的围岩变形动态预警指标及标准，实现了施工期动态安全预警。针对开挖扰动会导致围岩内部产生劣化损伤区的事实，本文引入边坡开挖中常用的强度折减法的计算思想，提出了基于实时数值仿真的施工期围岩劣化损失折减计算方法。该方法利用屈服接近度指标划定围岩劣化损失区的范围，每次只对劣化损伤区内的指定参数进行折减，直到数值计算不收敛为止，得到某施工状态下围岩变形与折减系数关系曲线；通过对曲线划分为匀速、加速和破坏三个变形过程，利用三段的分界值建立该开挖状态下的动态预警标准，进而实现了地下洞室群施工期的动态安全预警。（6）基于多层次模糊综合评价方法，提出了基于实测性态的地下洞室群安全性综合评价方法。本文基于对地下洞室群安全监测设计进行了深入分析与广泛调研，建立了基于实测信息的地下洞室群安全评价结构体系，采用层次分析法和专家打分法确定评价指标权重；然后通过混合使用―乘与取大算子M（·，∨）‖和―乘与和算子M（·，+）‖两种算子，对模糊综合评价方法的权重向量与模糊关系矩阵的合成进行改进，充分考虑断面最危险区域对整体安全性评价结果的重要性，将不稳定危险因素更好的向上层传递，使得最终的结构稳定性评判结果更加真实可信。此方法也为地下洞室群整体安全性评价提供一个新的思路。（7）综上，研发了基于物联网的地下洞室群施工专家信息系统，并对其它辅助功能进行简要展示。更多还原

【Abstract】 Underground cavern group construction is a complex system engineering,which is affected by a variety of complex natural geological conditions and manyother unknown factors. In view of the randomness, sudden, uncertainty and highlyempirical as well as other characteristics of the construction process of undergroundcavern group, the Internet of Things Technology (ITT) and other advancedtechnologies and methods are introduced in this paper. In addition, a systematicanalysis and research on the four main aspects (information collection, projectschedule, engineering safety, engineering warning) of the construction process ofunderground cavern group is carried out. Then an expert information system basedon ITT for the construction of underground cavern group is developed. The mainresearch contents and results of this paper are as follows:(1) Based on ITT, a real-time acquisition scheme for all kinds ofconstruction information of the underground cavern group is proposed and areal-time data acquisition platform for construction information of theunderground cavern group is built.ITT is first applied to the construction information collection of undergroundcavern group, with respect to the characteristics of various types of informationproduced in the construction process of underground cavern group; three differenttypes of information acquisition methods are proposed. Then information in theconstruction process can be collected timely and wirelessly. The problems of poortimeliness and highly complexity as well as other shortcomings of currentinformation gathering process in the construction of underground cavern group aremade up, and thus can provide real-time data for the scientific management andscientific decision in the course of construction of underground cavern group.(2)A time-varying construction schedule risk prediction method based onBayesian theory for construction of underground cavern group is put forwardin this paper; meanwhile, combined with the risk control concept ofconstruction schedule, a real-time risk control expert system for constructionschedule based on Bayesian theory is set up. Bayesian theory and real-time simulation technology are introduced into theprediction of underground cavern group construction schedule in this paper, atime-varying construction schedule risk prediction method based on Bayesiantheory for construction of underground cavern group is put forward. In the method,mean value and variance of unit workload construction time are chosen as randomvariables, and the probability distribution forecast of remaining work time of underconstruction process and work time of new construction procedure can be revisedand improved constantly according to the actual record information of theconstruction field. At the same time, combined with flexible network planningsimulation method, completion risk of the overall construction schedule of the lateproject can be predicted. With the method of this paper, construction scheduleprediction results are highly consistent with actual engineering practice, and thusthe method can play a practical role in promoting the development of probabilityprediction methods for the construction schedule of underground cavern group.(3)In this paper, a suitable method to establish numerical models forreal-time simulation is given and the real-time mapping relationship betweenactual construction state and numerical models is set up. Besides, a real-timenumerical simulation method based on numerical simulation technology forconstruction structural safety of underground cavern group is proposed.Through in-depth analysis of actual construction state and constructionnumerical simulation state, main factors of the two are given. Then by bringing inthe concepts of collection in mathematics, the real-time dynamic mapping relationsbetween the two states are established. The simulation of actual constructionprogress, unexpected adverse geological conditions and support measures oflarge-scale refinement numerical models can be completed timely andautomatically, and the aim of real-time numerical simulation for constructionstructural safety of underground cavern group is realized. Thus it can provideimportant reference for field engineers to make decisions and commands.(4)Integrated innovation point (2) and (3), the coupled real-time simulationmethod of structural safety for construction period of underground caverngroup is proposed. And based on structure safety and schedule coupledsimulation method, a treatment expert system for unexpected adversegeological conditions during construction is put forward. Construction safety and progress of large underground cavern group are greatlyinfluenced by sudden adverse geological conditions, thus real-time determination ofthe combined affects which sudden adverse geological conditions have onconstruction structure safety and schedule has important engineering significance.In this paper, sudden adverse geological conditions are mapped to numericalmodels. And based on the real-time numerical simulation results, structural safetyof the adverse geological areas is assessed. Then appropriate measures forconstruction states that don‘t meet the security requirements can be put forward.Through the real-time simulation of treatment effects, decided treatments in formsof construction parameters are sent back to the construction schedule control system,in this way, the system can assess the effects of treatments on overall constructionschedule. The system implements the on-site office integration of undergroundengineering construction safety and progress control and can provide technicalsupport for the security and progress control of the on-site construction ofunderground works.(5) Based on the idea of strength reduction and combine with the real-timenumerical simulation method mentioned in (3), a calculation method based onreal-time numerical simulation of the construction period of the undergroundcavern group surrounding rock deterioration damage reduction is proposed.Surrounding rock deformation dynamic early warning indicators andstandards of surrounding rock under certain construction status during theconstruction phase can be calculated through that method, and thus it canrealize the aim of dynamic security warning during the construction phase.In consideration of the fact that excavation disturbance can cause thepropagation of deterioration and damage zones in inner of surrounding rock, thispaper introduces the idea of strength reduction which is commonly used in theexcavation of slopes and proposes a calculation method based on real-timenumerical simulation of the construction period of the underground cavern groupsurrounding rock deterioration damage reduction. In this method, the range ofdeterioration damage zones of surrounding rock is divided by the yield proximityindicators. In each calculation only the strength parameters of deterioration damagezones are reduced, the strength of parameters should be reduced until the numericalcalculation does not converge, then a curve of surrounding rock deformation andthe reduction factor can be got; the curve can be divided into three deformation parts, namely the uniform part, the acceleration part and the destruction part.Boundary values of the three parts can be used to establish the dynamic earlywarning standards of that excavation status, and then the aim of dynamic securitywarning during the construction phase is realized.(6) Based on multi-level fuzzy comprehensive evaluation method, acomprehensive safety evaluation method according to observed behavior ofunderground cavern group is proposed.Based on in-depth analysis and extensive research on the security andmonitoring design of underground cavern group, a structure safety evaluationsystem of underground cavern group based on measured information is establishedin this article. Index weights in the system are determined through analytichierarchy process and expert scoring method. Then by comprehensive applicationof two kinds of operators,―the multiplication and max operator M(,∨)‖and―themultiplication and addition operator M(,+)‖, the weight vector of fuzzycomprehensive evaluation method and the composition of fuzzy relationship matrixcan be improved. With fully consideration of the influence of most dangeroussectional area on overall safety evaluation results, unstable risk factors can be betterdelivered to upper layer, thus the final evaluation result of structure stability is moreauthentic and believable. Furthermore, this method also provides a new way ofthinking for overall safety evaluation of underground cavern group.(7) In conclusion, based on ITT, a construction expert information systemof underground cavern group is developed and other auxiliary functions of thesystem are briefly introduced in the paper as well.更多还原