基于改进PSO-BP神经网络的浅埋隧道围岩变形时序预测研究

Time Series Prediction of Surrounding Rock Deformation in Shallow Buried Tunnel Based on Improved PSO-BP Neural Network

通过智能化的模式对浅埋隧道围岩变形进行时间序列预测研究,利用改进PSO算法优化BP神经网络参数的方法,用于预测浅埋隧道围岩变形数值。采用改进PSO-BP神经网络构建了海天堡浅埋隧道的拱顶沉降和周边收敛预测模型,使用主成分分析方法选取影响围岩变形的关键因素作为影响因子输入。通过对实验结果的比较和分析,实验结果显示改进PSO-BPNN算法的优越性。

In order to protect the personal safety of tunnel construction personnel,the time series prediction of surrounding rock deformation of shallow buried tunnel is carried out by using the intelligent model,and the improved PSO algorithm is used to optimize the parameters of BP neural network to predict the surrounding rock deformation value of shallow buried tunnel. The improved PSO-BP neural network is used to construct the prediction model of vault settlement and peripheral convergence of Haitianbao shallow buried tunnel in Chongqing. The principal component analysis method is used to select the key factors that affect the deformation of surrounding rock as the input influencing factors.Through the comparison and analysis of the experimental results,the experimental results show that the new model has a relatively higher accuracy.