Traffic Prediction Using Road Segment Travel Time Reliability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Mapping and navigation service providers face challenges in accurately predicting dynamic traffic conditions due to the difficulty in combining and blending different data sources and types effectively.
Innovation Solution
The method involves retrieving real-time and historical traffic information for each road segment, aggregating traffic flow speed data to compute traffic pattern data and travel time reliability index metrics, and determining the appropriate data to use for traffic prediction based on these metrics.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If traffic pattern data is used for traffic prediction after a certain time period, then the prediction process is simplified, but the accuracy of traffic prediction deteriorates
Solution Approach 1:
The system dynamically changes the parameter of data selection based on travel time reliability metrics. When reliability is high, it uses traffic pattern data; when reliability is low, it uses real-time traffic data. This parameter change resolves the contradiction by adapting the data source to current conditions, maintaining accuracy while allowing for simplified processing when appropriate.
Solution Approach 2:
The patent implements a dynamic blending approach where the weight assigned to traffic pattern data versus real-time traffic data changes based on computed reliability metrics. This dynamic adjustment allows the system to transition between using simplified historical patterns and more complex real-time data depending on current traffic conditions, resolving the trade-off between process simplicity and prediction accuracy.
2Productivity
If real-time traffic data and historical traffic data are blended using weight-assigning algorithm, then traffic prediction can be generated, but the quality and accuracy of prediction deteriorates due to ineffective blending
Solution Approach 1:
The system computes travel time reliability metrics as feedback to evaluate the quality of historical traffic data. This feedback mechanism determines the appropriate weighting between historical and real-time data, ensuring that blending only occurs when it improves prediction quality. The reliability metrics provide continuous feedback on data quality, enabling intelligent blending decisions that maintain high prediction accuracy.
Solution Approach 2:
The travel time reliability metric acts as an intermediary that mediates between historical traffic data and real-time traffic data. Rather than directly blending the two data sources, the system uses the reliability metric as an intermediate decision-making layer to determine the optimal blend, ensuring that the combination of data sources maintains high prediction quality while enabling continuous prediction generation.
3Adaptability or versatility
If traffic data blending is performed to enhance traffic prediction, then prediction coverage is improved, but the reliability of prediction deteriorates due to uncertainty in data quality
Solution Approach 1:
The system performs preliminary computation of travel time reliability metrics before conducting traffic data blending. This preliminary action assesses the quality and reliability of historical traffic data in advance, allowing the system to make informed decisions about whether and how to blend data sources. By evaluating data reliability beforehand, the system ensures that blending operations maintain prediction reliability while expanding prediction coverage across different traffic conditions.
Data Source
AI summary
An approach is provided for traffic data blending based on road segment travel time reliability during traffic prediction. The approach involves, for instance, retrieving real-time traffic information and/or historical traffic information for each road segment within a geographic area. The approach also involves aggregating traffic flow speed data in the real-time traffic information to compute traffic pattern data for each road segment. The traffic pattern data includes static speed data of each said road segment. The approach further involves aggregating traffic flow speed data in the historical traffic information to compute travel time reliability index metric(s) for each said road segment. The approach further involves determining, based on the travel time reliability index metric(s), to use the traffic pattern data, a mean, or a percentile of a road segment travel time distribution in the historical traffic information for traffic prediction associated with each said road segment.