TRAFFIC VIOLATIONS MODELING BASED ON BCD MODEL AND REDUNDANT PREDICTION TOOLS

Prediction of Highway Code violations is one of the key concepts that may improve safety by avoiding errors of drivers. For this, we analyse drivers' behaviour by the mean of several objective criteria relating to comfort and security. These criteria are taken as they are or are interpreted with the Benefits, Costs and potential Deficit (BCD) model before being computed by two different prediction methods. We show how the combination of methods improves the prediction rate as well as providing information about the results' reliability.