基于优化PSO-BP算法的轨道交通短期OD客流预测研究  

Research on Short-term OD Passenger Flow Forecast of Rail Transit Based on Optimized PSO-BP Algorithm

作　　者：宋丽梅 SONG Li-mei(Yangling Vocational and Technical College,Yangling,Shaanxi 712100,China)

机构地区：[1]杨凌职业技术学院,陕西杨凌712100

出　　处：《杨凌职业技术学院学报》2024年第2期21-23,59,共4页Journal of Yangling Vocational & Technical College

基　　金：杨凌职业技术学院2021年院内基金项目“城市轨道交通网络高峰客流拥挤管控研究”研究成果(ZK21-38)。

摘　　要：城市交通系统要实现更好的管理,需对城市轨道交通进站客流进行准确预测,为达到提高轨道交通运输效率、改善运营服务质量的目的,构建了以反向传播(BP)神经网络对地铁客流进行预测;利用PSO,对BP神经网络进行进一步优化,形成对应的客流预测系统。以地铁数据为基础,对车站OD客流量时空相关性进行定性分析,利用回归分析法对影响客流的因素进行定量分析,筛选出天气、节假日、运营时刻3个时间特征。为提高预测精度,构建不同时间段下的BP神经网络模型,优化了PSO-BP神经网络模型的预测误差,形成了基于PSO-BP神经网络的轨道交通短期OD客流量预测模型,加入时间特征的短期OD客流量预测模型,其换乘站优化后神经网络模型预测值M1平均下降了48.2%,M2下降了37.6%,M3下降了21.9%,该方法和模型为轨道交通运营部门制定列车运行计划提供更准确数据资料。To achieve better management of urban transportation systems,it is necessary to accurately predict the passenger flow of urban rail transit inbound.In order to achieve the purpose of improving rail transit efficiency and improving the quality of operation services,a back propagation(BP)neural network is constructed to predict the subway passenger flow;using the particle swarm optimization algorithm(PSO)to further optimize the BP neural network to form a subway passenger flow prediction system that considers the influence of complex factors.Based on the data of Xi’an Metro Line 1,this paper qualitatively analyzes the time-space correlation of station OD passenger flow,uses regression analysis to quantitatively analyze the factors affecting passenger flow,and screens out the three time characteristics of weather,holidays,and operating hours.In order to improve the prediction accuracy,this paper constructs the BP neural network model under different time periods,optimizes the prediction error of the PSO-BP neural network model,and forms a rail transit short-term OD passenger flow prediction model based on the PSO-BP neural network,adding time characteristics of the short-term OD passenger flow forecasting model.The PSO-BP neural network model prediction value M1 decreased by 48.2%on average,M2 decreased by 37.6%,and M3 decreased by 21.9%.This method and the model provides more accurate data for rail transit operation departments to formulate train operation plans.

关 键 词：城市轨道交通 BP神经网络 粒子群优化算法 回归分析法 OD客流量预测模型 

分 类 号：U293.13[交通运输工程—交通运输规划与管理]