摘要
城市轨道交通以其运载量大、安全、准时、环保等特点在我国迅速稳步发展,如何准确及时预测客运量成为热点问题。时间序列分析被运用在预测客流方面,外界发生重大事件,客流预测结果将会受到较大影响。为了有效弥补时间序列分析在客运量序列分析的不足,文章以成都一号线的客运量序列为例,将小波分析和时间序列模型结合,利用小波分解技术将原始时间序列进行小波分解,再对每层的相关系数进行时间序列预测,利用小波重构技术将各时间尺度的数据预报分析结果进行组合,成为地铁时间序列的最终预报,将小波分解序列的基本特性应用于时间序列分析。通过误差分析,文章提出的客流预测分析方具有一定的预测精度,且相较于其他传统分析方法预测精度更高。
Urban rail transit is developing rapidly and steadily in my country due to its large capacity,safety,punctuality,and environmental protection.How to accurately and timely predict passenger traffic has become a hot issue.Time series analysis is used to predict passenger flow.When a major event occurs outside,the result of passenger flow forecast will be greatly affected.In order to effectively make up for the insufficiency of time series analysis in passenger volume sequence analysis,the article takes the passenger volume sequence of Chengdu Line 1 as an example,combines wavelet analysis with time series model,and uses wavelet decomposition technology to decompose the original time series.The correlation coefficients of each layer are used to predict the time series.The wavelet reconstruction technology is used to combine the data forecast analysis results of various time scales to become the final forecast of the subway time series.The basic characteristics of the wavelet decomposition series are applied to the time series analysis.Through error analysis,the passenger flow forecasting analysis method proposed in the article has a certain forecasting accuracy,and the forecasting accuracy is higher than other traditional analysis methods.
作者
黄裕萌
朱云华
胡尧
周欣璐
HUANG Yu-meng;ZHU Yun-hua;HU Yao;ZHOU Xin-lu
出处
《智能城市》
2021年第18期128-130,共3页
Intelligent City
基金
四川省创新训练项目(201910615007)
西南石油大学大学生课外开放试验(KSZ19218)。
关键词
小波分析
时间序列
地铁客流预测
wavelet analysis
time series
subway passenger flow forecast
