| SELF-TUNING OCCUPANT INJURY PREDICTION ALGORITHM FOR ADVANCED AUTOMATIC CRASH NOTIFICATION SYSTEM
|
| Ying Lu 1, Yuezhou Yin 1, Heng Wei 2, Na Chen 3 |
1School of Automotive and Traffic Engineering, Jiangsu University
2Department of Civil & Architectural Engineering, University of Cincinnati
3School of Planning, College of Design, Architecture |
|
|
|
|
|
|
|
| ABSTRACT |
|
In this study, the effect of discrimination threshold on velocity variation is investigated, and a self-tuning
algorithm for velocity variation is proposed to improve the accuracy of an advanced automatic crash notification (AACN)
system. First, after determining the factors affecting driver injury, an injury prediction model for the driver is developed. Second,
the prediction accuracy affected by the discrimination threshold in the AACN system is analyzed using a finite element model
for a sled crash test. Third, the self-tuning occupant injury prediction algorithm is presented based on an association model for
the velocity variation error, discrimination threshold, and acceleration peak. Fourth, a vehicle terminal is designed by
embedding a self-tuning algorithm into the system. Finally, a sled test and vehicle crash test are conducted to verify the
reliability of the self-tuning algorithm. The test results show that the self-tuning algorithm can increase the accuracy of the
probability of driver injury. |
| Key Words:
Advanced automatic crash notification system, Discrimination threshold, Velocity variation, Self-tuning algorithm, Vehicle terminal
|
|
Corresponding Author.
Ying Lu , Email. 
Preview
Full text via DOI
Download Citation 
CrossRef TDM
Print 
