Title :
Neural network model for truck driver fatigue accident detection
Author :
Hamouda, Ghada ; Saccomanno, Frank F.
Author_Institution :
Dept. of Civil Eng., Waterloo Univ., Ont., Canada
Abstract :
Truck accidents account for a disproportionately high number of deaths and personal injuries on the highway network. Driver fatigue is perceived to be a major cause of these accidents. However, the presence of fatigue as documented in truck accident police reports does not support the commonly held perception that it is a problem. Either fatigue is not as much of a problem as perceived, or more likely, police accident reports significantly underestimate the involvement of driver fatigue in truck accidents. A neural network (NN) model provides a comprehensive method for identifying presence of fatigue in truck accidents from supplementary information in police report. A NN model is proposed to classify truck accident into fatigue and nonfatigue related. The model was able to learn the required function efficiently and a very high performance was observed
Keywords :
accidents; human factors; neural nets; traffic engineering computing; neural network model; police accident reports; truck driver fatigue accident detection; Biological neural networks; Civil engineering; Fatigue; Injuries; Neural networks; Road accidents; Road transportation; Statistical analysis; Telecommunication traffic; Traffic control;
Conference_Titel :
Electrical and Computer Engineering, 1995. Canadian Conference on
Conference_Location :
Montreal, Que.
Print_ISBN :
0-7803-2766-7
DOI :
10.1109/CCECE.1995.528150