基于B-P神经网络的北京地铁运量预测模型

Passenger Traffic Volume Forecast for Beijing Subway Based on Back Propagation ANN Model

分析影响北京地铁运量的因素,选取城市轨道交通内部生产、人口因素、乘客收入因素和交通因素等作为运量预测的关键要素,应用B-P神经网络(Back Propagation ANN,基于误差反向传播神经网络)模型,对未来几年内北京地铁运量进行预测分析。结果发现B-P神经网络对地铁客运量的预测较为适用,可为北京地铁运营规划提供决策参考。

Factors affecting passenger traffic volume in Beijing subway were analyzed.Selecting subway internal production,city population,passengers＇income and traffic as the key factors for the forecast,authors applied Back Propagation ANN model to predict and analyze the passenger traffic volume in the coming years in Beijing subway.Results indicated that B-P ANN model is adaptable to the passenger flow volume forecast for subways and can be adopted as the basis for the operation planning of Beijing subway.