京津塘高速公路K131+610段路基沉降预测研究

Study on the Prediction of Roadbed Settlement of K131+61 Section of Beijing-Tianjin-Tanggu Expressway

高速公路施工中,路基沉降对工程质量有着重要的影响。但是影响路基沉降的因素很多,传统的线性预测模型很难将各种影响因素统一表达,BP神经网络模型可以将传统的函数关系式转化为高维的非线性映射,在路基沉降预测中得到广泛应用。以京津塘高速公路K131+610段工程为例,探讨了BP神经网络模型的最佳参数配合。

In the construction of expressway, the roadbed settlement has important influence on the construction quality. However, there are many factors influencing the roadbed settlement, it is very difficult for the traditional liner prediction model to make the unified expression of various influencing factors, and the BP neural？network model, which can transform the functional relation expression into the high-dimensional nonlinear mapping, is widely used in the prediction of roadbed settlement. Taking the project of K131 ＋61 section of Beijing-Tianjin-Tanggu Expressway as an example, this paper probes into the optimum coordination of the parameters of BP neural network model.