摘要
为构建有效的小样本数据下的城市轨道交通客流预测模型,研究以西安地铁2号线历史客流量数据为基础,采用时间序列分析方法构建ARIMA与SARIMAX模型。通过数据变换、参数优化及节假日效应建模,实现客流量的高精度预测。结果表明,对数变换结合ARIMA(1,0,0)模型显著提升预测稳定性,数变换结合SARIMAX模型研究进一步验证了周季节性及节假日因素对客流波动的影响。对未来7天进行客流预测结果表明,预测结果满足精度要求。
In order to construct an effective passengerflow prediction model for urban rail transit based on small sample data,this study uses the time series analysis method to construct ARIMA and SARIMAX models based on the historical passenger flow data of Xi'an Metro Line 2.Through data transformation,parameter optimization and holiday effect modeling,high-precision prediction of passengerflow is realized.The results show that the logarithmic transformation combined with the ARIMA(1,0,0)model significantly improves the prediction stability,and the numerical transformation combined with the SARIMAX model further verifies the influence of weekly,seasonal and holiday factors on thefluctuation of passengerflow.The passengerflow forecast for the next 7 days shows that the prediction results meet the accuracy requirements.
作者
潘新丽
姚转香
王敬宇
Pan Xinli;Yao Zhuanxiang;Wang Jingyu
出处
《时代汽车》
2025年第15期187-189,共3页
Auto Time
基金
陕西省大学生创新创业训练项目“基于大数据的城市轨道交通客流预测模型研究”(项目编号:2024DC04)。
