Bayesian Travel Time Distribution Estimation on Signalized Arterials
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing systems for estimating traffic conditions on signalized arterials face challenges due to interruptions from traffic signals and sparse instrumentation, leading to low-quality data and significant variations in traffic estimations among service providers.
Innovation Solution
A system utilizing a processor and memory with an application that receives travel time data, applies Bayesian Inference principles to update prior distributions based on new observations, and calibrates estimates considering contextual factors like weather and events, using re-identification technologies and GPS-connected devices to improve data collection and accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional traffic detectors are used on signalized arterials, then traffic flow data can be collected, but instrumentation is sparse due to low traffic volumes making such instrumentation difficult to justify economically
Solution Approach 1:
The patent uses GPS-connected devices as intermediary elements to collect traffic data. These mobile GPS units act as temporary detectors that can be deployed flexibly without requiring permanent instrumentation infrastructure on the arterials, thus overcoming the economic justification barrier while still enabling data collection.
Solution Approach 2:
The system creates virtual copies of traffic detector functionality through software applications running on GPS-connected devices. Instead of deploying physical detectors, the system replicates detection capabilities using mobile computational devices that can report location and travel time data.
2Quantity of substance
If re-identification technologies are used to increase data collection, then sampling sizes increase significantly, but the quality of travel time estimates remains dubious due to limited observations and complex traffic patterns
Solution Approach 1:
The system implements feedback mechanisms where travel time estimates are continuously refined. Observed travel times from GPS devices are compared against predicted values, and the model parameters are adjusted based on this feedback to improve estimation accuracy over time, addressing the quality issue despite increased sampling.
Solution Approach 2:
The patent transforms the estimation problem by changing parameters from direct travel time measurement to probability distribution estimation. Instead of providing single point estimates, the system models travel times as probability distributions with varying parameters, allowing for more nuanced and accurate representations of uncertain traffic conditions.
3Productivity
If traffic signals are installed at intersections, then traffic flow control is improved, but traffic flows are interrupted generating complex traffic patterns that make estimation difficult
Solution Approach 1:
The patent segments the traffic analysis into multiple components: trip segments between intersections, signal cycle segments, and time-of-day segments. By breaking down the complex interrupted flow into manageable segments, the system can model each component separately and combine them to understand overall travel patterns.
Solution Approach 2:
The system employs dynamic modeling where traffic parameters are not fixed but vary over time and space. Travel time distributions are modeled as dynamic entities that change with traffic conditions, signal states, and temporal patterns, allowing the system to adapt to the complexity introduced by signalized intersections rather than treating them as static obstacles.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A system is provided for estimating time travel distributions on signalized arterials. The system may be implemented as a network service. Traffic data regarding a plurality of travel times on a signalized arterial may be received. A present distribution of the travel times on the signalized arterial may be determined. A prior distribution based on one or more travel time observations may also be determined. The present distribution may be calibrated based on the prior distribution.