隧道围岩变形动态预测的灰色自适应模型及其参数智能辨识  

Grey Self-adaptive Model of Dynamic Prediction of Surrounding Rock Deformation of Tunnel and Intelligent Identification of Parameters

作　　者：高文华[1] 朱建群[1] 黄自永[1] 邓修甫[1] 

机构地区：[1]湖南科技大学岩土工程研究所,湖南湘潭411201

出　　处：《公路交通科技》2012年第1期114-120,149,共8页Journal of Highway and Transportation Research and Development

基　　金：国家自然科学基金项目(50874043);湖南省教育厅重点科研项目(09A028);湖南省科技计划项目(2010FJ3046)

摘　　要：隧道围岩变形具有动态性、对时间和空间的敏感性及高度的非线性等特征。基于这些特性,探讨了隧道围岩变形动态预测灰色自适应GM(1,1)模型的建模思路及步骤。为确保参数为全局最优,提高模型的预测精度,提出采用粒子群优化算法(PSO)对模型参数进行智能辨识,并给出了参数智能辨识的方法和步骤。根据所建立的模型,编制了相应的计算程序。在此基础上,以贵州厦蓉高速公路某施工标段为工程实例,建立了隧道拱顶沉降及收敛位移预测分析的灰色自适应模型。算例分析表明,在数据较少的情况下,基于PSO参数智能辨识的灰色自适应模型的递推和跟踪能力较强,预测结果与实际监控数据吻合程度较高,是隧道围岩变形动态预测较为有效的方法,可在实际工程中推广应用。The surrounding rock deformation of tunnel has the characteristics of dynamics, sensibility to time and space, and high nonlinearity. According to these characteristics, the idea and process of setting up the grey self-adaptive model GM （ 1,1 ） for dynamic prediction of surrounding rock deformation of tunnel was discussed. To ensure that the parameters gained are globally optimal solution, the particle swarm optimization （PSO） was used for intelligent identification of the model parameters so as to improve prediction precision of the model, and the method and procedure for intelligent identification of parameters were given. The computing program was compiled according to the established model. Taken a construction section of Xiamen- Chengdu expressway in Guizhou for example, the grey self-adaptive models for predicting arch crown settlement and convergence displacement of tunnel were set up. The example analysis shows that the grey adaptive model based on PSO intelligent identification of parameters has the better recurrence and tracking ability with insufficient data, and the prediction results are in good accordance with the monitoring data insitu. The suggested method is effective to dynamic prediction of surrounding rock deformation of tunnel, and could be popularized and applied in actual engineering.

关 键 词：隧道工程 围岩变形 灰色自适应模型 粒子群优化算法 动态预测 

分 类 号：U456[建筑科学—桥梁与隧道工程] TU454[交通运输工程—道路与铁道工程]