【摘要】 分析了影响道面使用性能的各种参数,结合BP神经网络和遗传算法来预测机场道面使用性能。通过遗传算法全局寻优功能对神经网络的初始权值和阈值进行优化,然后采用LM(Levenberg-Marquardt)优化算法对神经网络训练速度进行加速,并且使训练避免陷入局部极小点。通过历年数据对神经网络进行训练,用所得神经网络模型对机场道面使用性能进行预测。训练结果表明,该方法具有足够的精度,能够应用到工程实际中。更多还原

【Abstract】 The problem of forecasting the service performance of airport pavement is studied.Various parameters influencing the service performance of airport pavement are analyzed,the genetic algorithm and BP neural network are combined to forecast the service performance of airport pavement.The genetic algorithm overall optimization is used to optimize the weight and threshold of neural network,then,LM optimization algorithm is used to increase the training speed and make the training avoid getting into local minimum.The neural network is trained with the data of the past years,and the model of neural network is used to forecast the service performance of airport pavement.The result shows that this method is precise enough to be applied to the engineering practice.更多还原