The invention discloses a large-scale short-time traffic flow prediction method based on a progressive mode, and the method comprises the steps: S1, obtaining an original data set collected by a sensor, and carrying out the preprocessing of the original data set; s2, constructing a large-scale short-time traffic flow model based on a progressive mode, and training the large-scale short-time traffic flow model by taking the processed multiple periodic time sequence data and the driving distance adjacent matrix as the input of the model; and S3, inputting the current traffic data through the obtained model to predict future traffic data. According to the method, based on a progressive thought, time features and space features are fused step by step to carry out time-space fusion operation more comprehensively and fully, so that the problem of insufficient time-space fusion in large-scale traffic flow prediction of an existing method is effectively solved; the time overhead of calculation and spatial dependence modeling is greatly reduced; and meanwhile, the prediction effect is more accurate, and a more accurate reference basis can be provided for urban traffic management.
本发明公开了一种基于渐进式的大规模短时交通流量预测方法,包括:S1、获取传感器采集的原始数据集,并对原始数据集进行预处理;S2、构建基于渐进式的大规模短时交通流量模型,将处理好的多种周期时间序列数据和驾车距离邻接矩阵作为模型的输入,训练所述大规模短时交通流量模型;S3、通过得到的模型对其输入当下的流量数据来预测未来的流量数据。该方法基于渐进式的思想,将时间特征和空间特征通过逐步融合的方式来更全面、更充分地进行时空融合操作,有效地解决现有方法在大规模交通流量预测中存在的时空融合不充分的问题,且大幅度降低了计算和空间依赖建模的时间开销;同时,预测效果更精准,能够为城市交通管理提供较为准确的参考依据。
Progressive-based large-scale short-term traffic flow prediction method
基于渐进式的大规模短时交通流量预测方法
2023-07-04
Patent
Electronic Resource
Chinese
IPC: | G08G Anlagen zur Steuerung, Regelung oder Überwachung des Verkehrs , TRAFFIC CONTROL SYSTEMS / G06F ELECTRIC DIGITAL DATA PROCESSING , Elektrische digitale Datenverarbeitung / G06N COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS , Rechnersysteme, basierend auf spezifischen Rechenmodellen / G06Q Datenverarbeitungssysteme oder -verfahren, besonders angepasst an verwaltungstechnische, geschäftliche, finanzielle oder betriebswirtschaftliche Zwecke, sowie an geschäftsbezogene Überwachungs- oder Voraussagezwecke , DATA PROCESSING SYSTEMS OR METHODS, SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL, SUPERVISORY OR FORECASTING PURPOSES |
British Library Conference Proceedings | 2020
|Event-Based Short-Term Traffic Flow Prediction Model
British Library Conference Proceedings | 1995
|