Related Experiment Videos
An algorithm for assessing the risk of traffic accident.
Kwok-Suen Ng1, Wing-Tat Hung, Wing-Gun Wong
1Department of Civil and Structural Engineering, Hong Kong Polytechnic University, Hung Hom, Kowloon, Hong Kong, China. ceks.ng@polyu.edu.hk
Journal of Safety Research
|October 31, 2002
Summary
This study developed an algorithm combining GIS and statistical methods to improve traffic accident risk estimation. The new method enhances accuracy, particularly for fatal and pedestrian-related incidents, aiding authorities in identifying high-risk areas.
Area of Science:
- Traffic safety analysis
- Geospatial data science
- Statistical modeling
Background:
- Developing accurate methods for estimating traffic accident frequency and risk is crucial for public safety.
- Existing methods may not fully capture the complex factors influencing accident occurrence.
Purpose of the Study:
- To create a novel algorithm for estimating traffic accident numbers and assessing accident risk.
- To integrate Geographical Information System (GIS) techniques with statistical analysis for enhanced prediction.
Main Methods:
- Utilized GIS for spatial mapping and distribution analysis of accidents.
- Employed cluster analysis for data grouping and regression analysis (Negative Binomial) to model accident factors.
- Incorporated the Empirical Bayes (EB) approach for robust accident risk computation.
Main Results:
- The developed algorithm significantly improved accident risk estimation compared to traditional methods relying solely on historical data.
- Demonstrated enhanced efficiency in analyzing fatality and pedestrian-involved traffic accidents.
Conclusions:
- The algorithm provides valuable insights for authorities to pinpoint areas with elevated accident risk.
- Offers a data-driven reference for urban planners and policymakers to enhance road safety strategies.