Traffic State Prediction Using Link Section Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional techniques fail to accurately predict the detailed congestion states of a link's sections, as they treat the entire link uniformly, ignoring variations such as intersections or slopes, which can cause congestion to differ along the link.
Innovation Solution
A traffic state predicting apparatus that classifies link sections based on their congestion levels using predicted travel time and specific parameters like smooth and congested traffic speeds, and congestion reference positions, enabling more precise congestion state prediction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If the entire link is treated uniformly to obtain travel time, then the measurement process is simple, but the congestion state prediction precision is insufficient
Solution Approach 1:
The link is divided into multiple sections based on congestion reference positions (such as intersections, slopes, or other bottleneck locations). Each section is independently analyzed to determine its congestion state, allowing for more precise localization of congestion patterns while maintaining a manageable prediction framework.
Solution Approach 2:
Different sections of the link are assigned different congestion characteristics based on their specific features. By using congestion reference positions to identify unique sections, the system applies local quality analysis to predict congestion states more accurately for each segment rather than treating the entire link uniformly.
2Loss of information
If detailed section classification is implemented, then the congestion state information becomes more informative, but the data processing complexity increases
Solution Approach 1:
Congestion reference positions are pre-established based on known bottleneck locations such as intersections, slopes, or other fixed infrastructure features. This preliminary classification of reference positions enables subsequent real-time prediction to focus only on determining whether congestion occurs in each predefined section, rather than analyzing every possible location.
Solution Approach 2:
The system uses a standardized model for section classification that can be replicated across different links. By creating a template based on congestion reference positions that can be applied uniformly, the system reduces processing complexity while maintaining detailed section-level information across multiple links.
Data Source
AI summary
A traffic state predicting apparatus, which predicts congestion states of parts of a link based on limited information (the travel time for the link), comprises storage means that stores link data including a link length of each link forming part of a road on a map and parameters including, for each link, a smooth traffic speed indicting being smooth, a congested traffic speed indicating being congested, and a congestion reference position which is a reference position for a congested section; means that acquires a predicted travel time for the link; and congestion degree classified section calculating means that obtains sections classified according to their level of congestion in the link with use of the predicted travel time and the parameters.