基于BP神经网络的铁路客运站交通接驳研究

Research on Traffic Connection of Railway Passenger Station Based on BP Neural Network

为优化客运站旅客接驳选择,在详细分析客运站旅客接驳流程的基础上,创建交通接驳指标体系。以西安北站为例,选取步行长度、接驳费用、接驳耗时、接驳时间、携带行李和接驳目的为输入指标,以接驳选择,如地铁、公交、出租车和网约车等为输出,构建BP神经网络。其次通过问卷调查、手机软件和资料查询的方法,量化各输入指标;根据旅客接驳选择情况,确定输出指标所占权重。经过训练后,BP神经网络训练数据正确率达到0.8542,测试数据正确率达到0.8380。最后任意选取2组输入指标验证神经网络的分类情况,结果证明:输入指标与输出指标有较为强烈的非线性关系,BP神经网络对优化旅客接驳选择,提升客运组织效率具有较为积极的影响。

In order to optimize the selection of passenger connection in passenger station,the traffic connection index system is established based on the detailed analysis of passenger connection process in passenger station.Firstly,taking Xi'an North Railway Station as an example,the BP neural network is constructed by selecting walking length,connection cost,connection time,baggage and connection purpose as input indexes,and connecting options such as subway,bus,taxi and online car as output.Secondly,by means of questionnaire survey,mobile phone software and data query,each input index is quantified.According to the passenger connection selection,the weight of the output index is determined.After training,the correct rate of BP neural network training data reached 0.8542,and the correct rate of test data reached 0.8380.Finally,2 groups of input indicators are randomly selected to verify the classification of the neural network.The results show that there is a strong nonlinear relationship between the input indicators and the output indicators,and BP neural network has a positive impact on optimizing passenger connection selection and improving the efficiency of passenger transportation organization.