The invention discloses a traffic flow time sequence decomposition method. The method comprises the following steps: (1) collecting traffic flow historical data of a road section to be analyzed; (2) decomposing a historical traffic flow sequence into a trend component, a period component and a residual component; and (3) decomposing a newly observed traffic flow sequence into three components of trend, period and residual error. According to the method, the periodicity of the traffic flow sequences is fully considered, the influence of random noise can be stripped, and traffic flow periodic fluctuation and trend change rules are extracted. The method is suitable for a traffic flow state sequence with periodicity, and a decomposition result can be used for traffic state analysis, detectionand prediction and the auxiliary research of an intelligent traffic system.
本发明公开了一种交通流时间序列分解方法,它包括以下步骤:(1)采集所需要分析道路断面的交通流历史数据;(2)将历史交通流序列分解为趋势、周期和残差3个成分;(3)将新观测的交通流序列分解为趋势、周期和残差3个成分。该方法充分考虑了交通流序列的周期性,可以剥离出随机噪声的影响,提取交通流周期波动以及趋势变化的规律。该方法适用于具有周期性的交通流状态序列,分解结果可用于交通状态分析、检测与预测,辅助智能交通系统的研究。
Traffic flow time series decomposition method
一种交通流时间序列分解方法
2020-05-19
Patent
Elektronische Ressource
Chinesisch
IPC: | G08G Anlagen zur Steuerung, Regelung oder Überwachung des Verkehrs , TRAFFIC CONTROL SYSTEMS |
Short-term traffic flow prediction method based on time series decomposition
Europäisches Patentamt | 2020
|Traffic flow time series prediction method and device
Europäisches Patentamt | 2020
|Adaptable Anomaly Detection in Traffic Flow Time Series
IEEE | 2019
|TRAFFIC FLOW TIME SERIES PREDICTION BASED ON STATISTICS LEARNING THEORY
British Library Conference Proceedings | 2002
|