TY  - JOUR
PY  - 2017//
TI  - Development of logistical regression models to predict rider injury severity for motorcycle-to-passenger car crashes
JO  - Transactions of Society of Automotive Engineers of Japan
A1  - Nishimoto, Tetsuya
A1  - Mukaigawa, Kosuke
SP  - 103
EP  - 109
VL  - 48
IS  - 1
N2  - In previous research, we developed an algorithm for an Advanced Automatic Collision Notification (AACN) system that predicted the injury severity resulting from car and bicycle accidents and one for car and pedestrian accidents. The purpose of the present study was to develop an injury prediction algorithm for car and motorcycle crashes. Two algorithms were developed using macro data from the 880,000 motorcycle (MC) accidents in the ITARDA database that occurred between from 2005 to 2014 in Japan. The first algorithm, MC model 1 uses crash details such as motorcycle travel speed and crash direction as well as motorcyclist information such as age of rider and helmet usage. The second algorithm, MC model 2 also uses the partner crash car information such as travel speed and crash direction. MC model 2, which uses both motorcycle and car travel speed information is a better prediction algorithm. The results indicate that for a 10% rate of under triage, the threshold is 6.4% for MC model 2.<p />  <p>Language: ja</p>
LA  - ja
SN  - 0287-8321
UR  - http://dx.doi.org/10.11351/jsaeronbun.48.103
ID  - ref1
ER  -