RES2021-09: Using Big Data and Machine Learning to Evaluate and Optimize the Performance of Traffic Signals in Tennessee
Transportation agencies are tasked with making decisions and taking actions to fight congestion and improve road conditions using the methods and resources available to them. Performance evaluations of traffic conditions at signalized intersections and arterials provide practical data for the agencies to make well-informed decisions for targeted retiming solutions. However, there are no widely accepted standards of evaluating traffic signals, and examining each intersection manually is very costly and time consuming. As a result, many agencies simply retime the traffic signals every 3-5 years or rely on citizens’ complaints to prioritize traffic signal retiming. In this work, segmented probe vehicle data from the RITIS website was used to implement a cost-effective approach that ranks the major traffic signals in Tennessee in the scale of 0 to 10. First, signalized intersections were extracted using the TMC segments. Then, three metrics were selected based on the available data from the RITIS website, and a ranking formula that incorporate these metrics as well as factors such as different time of the day and different days of the week was designed. September 2021 traffic data was used to calculate the ranking of the intersections, and an online database was developed to display, browse, and query the traffic signal ranking information. Kmeans, an unsupervised machine learning approach, was utilized to divide the signals into 6 clusters, using which the weighting factors of the ranking formula were finetuned. This work is the first step towards an automated evaluation system that can monitor the performance of traffic signals in real-time.
- Record URL:
- Record URL:
Language
- English
Project
- Status: Completed
- Funding: $ 104707
-
Contract Numbers:
RES2021-09
-
Sponsor Organizations:
Tennessee Department of Transportation
James K. Polk Building
Fifth and Deaderick Street
Nashville, TN United States 37243-0349 -
Project Managers:
Bryan, Stephen
-
Performing Organizations:
Middle Tennessee State University, Murfreesboro
1301 East Main Street
Murfreesboro, TN United States 37132-0001 -
Principal Investigators:
Miao, Lei
- Start Date: 20200901
- Expected Completion Date: 20220831
- Actual Completion Date: 0
- USDOT Program: Transportation, Planning, Research, and Development
Subject/Index Terms
- TRT Terms: Data analysis; Evaluation and assessment; Machine learning; Optimization; Performance measurement; Traffic signals
- Geographic Terms: Tennessee
- Subject Areas: Data and Information Technology; Highways; Operations and Traffic Management;
Filing Info
- Accession Number: 01754305
- Record Type: Research project
- Source Agency: Tennessee Department of Transportation
- Contract Numbers: RES2021-09
- Files: RIP, STATEDOT
- Created Date: Oct 7 2020 12:48PM