基于GA-BP神经网络的运营高速公路隧道养护工程科学决策优先级排序

Scientific Decision Prioritization of Operational Expressway Tunnel Maintenance Engineering based on GA-BP Neural Network

为解决运营高速公路隧道养护工程科学决策优先级排序计算量大,指标权重确定主观性、经验化等问题,基于GA-BP神经网络建立了运营隧道养护工程科学决策优先级排序模型。首先选取交通量、降水量、斜井竖井、不良地质状况、车道数量、通车年限、隧道长度和土建技术状况评分等8个特征指标,建立了量化制、评分制的运营高速公路隧道健康度评价标准。并通过遗传算法优化BP神经网络的初始权值和阈值,解决BP神经网络训练出现局部最优解的缺陷。最后,选取701个样本进行模型的训练和预测,结果表明,所建立的模型能快速准确计算隧道健康度评分,实现运营高速公路隧道养护工程科学决策优先级排序的客观化、科学化和智能化。

In order to solve the problems of heavy-load computation,subjective and empirical determination of index weights,a scientific decision prioritization model for operational tunnel maintenance engineering is established based on GA-BP neural network.Firstly,eight characteristic indexes of traffic volume,precipitation,inclined shaft,unfavorable geological conditions,number of lanes,in-service years,tunnel length and civil construction technology status are selected to establish the health evaluation standard of in-service expressway tunnel with quantitative and scoring system.The initial weights and thresholds of BP neural network are optimized by genetic algorithm to solve the defects of local optimal solution in BP neural network training.Finally,701 samples are selected for model training and prediction.The results show that the established model can quickly and accurately calculate the tunnel health score,and realize the objective,scientific and intelligent prioritization of scientific decision making in operational expressway tunnel maintenance.