摘要
以赣定高速公路谷山隧道工程为例,利用遗传算法和BP神经网络相结合的方法,建立运营隧道结构安全评估模型。网络结构采用遗传算法优化BP神经网络权值和阈值,提高了网络收敛速度,克服了传统的神经网络训练时间长、容易陷入局部极小值的问题。并将该模型应用到谷山隧道运营结构安全性评估中,具有较高的学习精度和较快的收敛速度。
Taking Ganding Highway Gushan Tunnel engineering for example, the evaluation model for operating tunnel structural safety has been established using genetic algorithm combined with BP neural network. This network structure adopts genetic algorithm to optimize the weights and threshold of BP neural network, which improved network convergence rate and overcame traditional problems that neural network will be trained for a long time and easily involved in local minimum. This model has been also applied in evaluation for Gushan Tunnel structural safety, which has higher learning accuracy and faster convergence rate.
出处
《路基工程》
2011年第6期111-114,共4页
Subgrade Engineering
基金
南昌航空大学研究生创新项目(YC2008032)
关键词
遗传算法
BP神经网络
运营隧道
结构安全
genetic algorithm
BP neural network
operating tunnel
structural safety