Abstract
This paper presents development and application of a statistical crash prediction model for various types of crashes in Chinese roadway segments. The model is constructed based upon a Negative Binomial Generalized Linear Model and is applied for a large amount of data collected from a wide range of urban, suburban and rural areas. The Negative Binomial Regression proposes a link function to fit a set of roadway characteristics data and traffic flow with crash frequency and at the same time handles the overdispersion problem. Through a real-world example, the performance of the model is evaluated and practical issues regarding input data quality issues and model validation are discussed. The results reveal that the proposed model can appropriately predict the crash data and enables safety traffic engineers to identify and prioritize the high crash locations and diagnose the roadway characteristics, which significantly affect the crash frequencies.
| Original language | English (US) |
|---|---|
| State | Published - 2014 |
| Event | 21st World Congress on Intelligent Transport Systems: Reinventing Transportation in Our Connected World, ITSWC 2014 - Detroit, United States Duration: Sep 7 2014 → Sep 11 2014 |
Other
| Other | 21st World Congress on Intelligent Transport Systems: Reinventing Transportation in Our Connected World, ITSWC 2014 |
|---|---|
| Country/Territory | United States |
| City | Detroit |
| Period | 9/7/14 → 9/11/14 |
All Science Journal Classification (ASJC) codes
- Control and Systems Engineering
- Mechanical Engineering
- Automotive Engineering
- Transportation
- Electrical and Electronic Engineering
Keywords
- Calibration
- Empirical Bayesian
- Negative Binomial Regression
- Safety performance function