论文部分内容阅读
路基沉降计算和预测是工程建设中一个重要的课题,但是很多情况下理论计算的沉降量与实际沉降量有较大的出入。所以必须对路基沉降进行监测,根据实测数据控制填土速率,保证路堤在施工中的安全与稳定;根据实测沉降曲线预测工后沉降,使工后沉降控制在设计允许范围之内。但沉降预测模型的参数是不确定的,用概率和概率分布去描述更加合适,将模型参数视为随机变量,基于贝叶斯理论和MCMC方法,借助WinBUGS软件,建立了贝叶斯沉降时间序列不确定性预测模型。实例分析表明,该方法所得结果合理可信,将其应用于路基沉降预测是可行和值得研究的。
Subgrade settlement calculation and prediction is an important issue in engineering construction, but in many cases there is a large discrepancy between theoretical settlement and actual settlement. Therefore, it is necessary to monitor the subgrade settlement and control the fill rate according to the measured data so as to ensure the safety and stability of the embankment during construction. The post-construction settlement is predicted based on the measured settlement curve so that post-construction settlement control is within the design allowable range. However, the parameters of the subsidence prediction model are uncertain. Probability and probability distributions are more suitable for description. The model parameters are regarded as random variables. Based on Bayesian theory and MCMC, WinBUGS software is used to establish the Bayesian subsidence time series Uncertainty prediction model. The case study shows that the result of the method is reasonable and credible, and it is feasible and worth to be applied to the subgrade settlement prediction.