Abstract
IHSDM (interactive highway safety design model) was used to predict accidents of 11 types of freeways, and the predicted accident numbers were compared with the actual accident numbers. The basic accident prediction models and the modification coefficients of alignment indexes were obtained, a set of IHSDM accident prediction improved models were built to carry out the sensitivity analysis of alignment indexes, the effects of alignment indexes on the predicted accident rates were explained, and the safe value ranges of alignment indexes for freeways were obtained. Analysis result shows that the accident numbers predicted by IHSDM are lower than the actual accident number, the least error is 13%, and the maximum error is 52%. The relative error between total accident number predicted by the improved model and the actual value is less than 10%. At specific road sections, Pearson's correlation coefficients are 0.60 at least, the root mean square errors are less than 0.30, and the mean relative errors are less than 30%. The sensitivities of accident rates to alignment indexes from high to low are straight line length, longitudinal slope, horizontal curve angle, horizontal curve radius and vertical curve radius. Obviously, compared with IHSDM, the improved model has higher accuracy and better applicability.
| Original language | English |
|---|---|
| Pages (from-to) | 123-132 |
| Number of pages | 10 |
| Journal | Jiaotong Yunshu Gongcheng Xuebao/Journal of Traffic and Transportation Engineering |
| Volume | 16 |
| Issue number | 1 |
| State | Published - 1 Feb 2016 |
| Externally published | Yes |
Keywords
- Accident prediction model
- Freeway
- IHSDM
- Sensitivity analysis
- Traffic safety
Fingerprint
Dive into the research topics of 'IHSDM freeway accident prediction models'. Together they form a unique fingerprint.Cite this
- APA
- Author
- BIBTEX
- Harvard
- Standard
- RIS
- Vancouver