According to statistics, articulated vehicles like tractor-semitrailers represent over 55% of heavy goods traffic in France. In some cases, this type of truck may have unstable behavior, resulting in high risks of accidents. Jackknifing is defined by a loss of yaw stability of the articulated system. It most commonly occurs when the semi-trailer is empty. This paper presents the development of an intelligent jackknifing safety system. The objective is to develop a strategy to detect a critical situation (jackknifing) and trigger a warning message under dangerous conditions such unstable yaw. A new system for articulated vehicles jackknifing detection and prediction is designed. This system uses on one hand, a nominal model of articulated vehicle making it possible to determine the vehicle dynamic state such as the relative yaw angle. On the other hand, detection algorithm is based on a jackknifing criterion and on the prediction function of jackknifing, in view of estimating the time to jackknifing. The validation of the results is done by using PROSPER, which is a very detailed Commercial simulator.
