摘要
城市交通拥堵情况预测序列多为单层级处理,导致预测结果误差较大,为此提出基于时间序列模型的城市交通拥堵情况预测方法。首先,进行预测指标选取及数据采集,采用多目标的方式扩大预测范围;其次,设计多目标预测序列,进行交通拥堵状态判别及无量纲化处理;最后,构建时间序列交通拥堵情况预测模型,并进行对比分析。测试结果表明,该预测方法的针对性较强,能够将预测绝对误差控制在1.5以下,具有实际应用价值。
The prediction sequence of urban traffic congestion is mostly processed at a single level,resulting in significant errors in the prediction results.Therefore,a time series model based method for predicting urban traffic congestion is proposed.Firstly,select prediction indicators and collect data,using a multi-objective approach to expand the prediction range.Secondly,design a multi-objective prediction sequence for traffic congestion state discrimination and dimensionless processing.Finally,construct a time series traffic congestion prediction model and conduct comparative analysis.The test results show that the prediction method has strong pertinence and can control the absolute error of prediction below 1.5,which has practical application value.
作者
任山山
买欣蕾
REN Shanshan;MAI Xinei(Department of Basic Education,Huanghe Jiaotong University,Jiaozuo Henan 454950,China)
出处
《信息与电脑》
2023年第10期1-3,共3页
Information & Computer
关键词
时间序列
序列模型
城市交通
拥堵情况
预测方法
time series
sequence model
urban transportation
congestion situation
prediction methods