基于GA-BP神经网络的航空货运量预测  

Air Cargo Volume Forecasting Based on GA-BP Neural Network

作　　者：徐森雨 田利军[1] XU Senyu;TIAN Lijun(Civil Aviation College,Nanjing University of Aeronautics and Astronautics,Nanjing 210016,China;Xinjiang Air Traffic Management Bureau CAAC,Urumqi 830016,China)

机构地区：[1]中国民航大学交通科学与工程学院,天津300300

出　　处：《综合运输》2023年第9期75-79,共5页China Transportation Review

基　　金：天津市教委社会科学重大项目(2021JWZD39)。

摘　　要：近年来,我国航空货运发展迅速,货运量实现连年增长,为把握未来航空货运发展趋势,本文利用人工神经网络学习能力强、非线性预测等优点,选取2005年-2022年我国航空货运量历史数据作为样本,构建了神经网络预测模型。在传统BP神经网络基础上,为解决网络初始权重和阈值随机赋值,导致其易陷入局部最小值的问题,通过利用遗传算法全局搜索和全局优化能力强的特性,建立遗传算法优化的神经网络预测模型,对2020年12月至2022年7月的货运量进行预测并和传统神经网络进行比较,实例证明GA-BP神经网络预测模型预测效果较好,提高了网络模型的预测精度,实现了更为精准的货运量预测。In recent years,China's air cargo has developed rapidly,and the cargo volume has achieved year after year growth.To grasp the future development trend of air cargo,this paper used the advantages of artificial neural network with strong learning ability and nonlinear prediction,selected the historical data of China's air cargo volume from 2005 to 2022 as samples,and constructed a neural network prediction model.Based on the traditional BP neural network,in order to solve the problem that the initial weights and thresholds of the network are assigned randomly,which causes it to fall into local minima,a neural network prediction model optimized by genetic algorithm,which is established by using the characteristics of strong global search and global optimization ability of genetic algorithm to predict the cargo volume from December 2020 to July 2022.Besides,through compared it with the traditional neural network,the example proved the GA-BP neural network prediction model has a better prediction effect,improved the prediction accuracy of the network model,and achieved more accurate freight volume prediction.

关 键 词：BP神经网络 遗传算法 优化 航空货运量 预测模型 

分 类 号：F562[经济管理—产业经济]