The invention relates to a road network heavy truck traffic flow prediction method based on data quality control, which is characterized in that two data sources of GPS flow data and toll station monitoring data are divided into three types of data road sections, and different prediction methods are respectively adopted. For a road section with GPS flow data, due to the fact that the GPS data arenot completely obtained, a method that sample expansion is conducted through a piecewise constant coefficient method and then long-short-term recurrent neural network prediction is conducted is adopted. And for a data-free road section without GPS data, a K nearest neighbor algorithm is adopted to predict the flow. For a toll station monitoring data road section, due to comprehensive data samples,a long-short-term recurrent neural network algorithm is directly adopted for prediction. According to the method, starting from the actual engineering problem faced by flow prediction, multiple datasource characteristics are analyzed to improve the data quality, then the road network heavy truck flow measurement and calculation method is established, and finally the road network heavy truck flowbased on data quality control is formed.
本发明涉及一种基于数据质量控制的道路网重型货车交通流量预测方法,针对GPS流量数据及收费站监测数据两种数据源划分三类数据路段,分别采取不同的预测方法。针对有GPS流量数据的路段,由于GPS数据存在获取不够全样问题,采用先分段常系数法扩样再长短时循环神经网络预测的方法。针对无GPS数据的无数据路段,采用K最近邻算法预测流量。针对收费站监测数据路段,由于数据样本全面,直接采取长短时循环神经网络算法进行预测。本发明从流量预测面对的实际工程问题出发,分析多种数据来源特征来提高数据质量,进而建立路网重型货车流量测算方法,最终形成基于数据质量控制的道路网重型货车流量。
Road network heavy truck traffic flow prediction method based on data quality control
基于数据质量控制的道路网重型货车交通流量预测方法
2020-08-25
Patent
Electronic Resource
Chinese
Road network dynamic traffic distribution method based on traffic flow prediction
European Patent Office | 2024
|Road network flow prediction method for urban traffic signal control
European Patent Office | 2023
|Heavy-duty truck fuel economy test in actual road traffic
Automotive engineering | 1997
|Heavy-Duty Truck Fuel Economy Test in Actual Road Traffic
SAE Technical Papers | 1997
|Truck traffic flow prediction method based on K clustering algorithm and neural network
European Patent Office | 2023
|