Prediction of traffic accident severity based on fuzzy logic

Prediction of traffic accident severity is one of the most important issues in traffic safety. Identification of effective characteristics in traffic accidents and their efficiency in severeness of crashes makes engineers able for further emprise which can reduce the number and probability of car accidents. Various parameters affect this prediction, theses parameters are vehicle speed, driver’s age, driver’ gender, type of vehicle, safety of vehicle, type of collision, seat belt use, point of impact, weather condition, traffic flow, behaviors of drivers. So eleven parameters and three injury severity levels are selected as input and output variables. In this paper, the effectiveness of the parameters and level of injury severity are predicted by using fuzzy logic. Traffic accident data of freeways were collected as train data in fuzzy system. Modeling result showed that the prediction accuracy was 88/3% for fuzzy logic. These results indicate a more accurate prediction ability of injury severity for fuzzy logic over other traditional methods. Results of the fuzzy logic indicated that driver’s age, speed, and seat belt, type of vehicle affect chances of experiencing a severe injury.

کلیدواژه ها:

Logic ، Traffic accident ، Prediction