Traffic monitoring and control are concerned by the knowledge of the compliance rate of the drivers with the road signs, and by the means to increase it, for instance, to a certain extent, the accumulation of redundant signs. A site presenting a set of successive road signs allows the estimation of the compliance rate evolution. Some statistical methods are available, if an adequate collection of data is performed. They are based on Before / After comparisons, in the time and / or in the space. Some problems are solved, as the recurrent impact of a driver on his follower; it appears as "auto correlation" from a statistical point of view; it can be processed by considering "platoons" of vehicles. Yet, some problems remain, mainly because, on the field, the drivers samples are not really known neither controlled. In this communication, these methods are presented and applied on the case of speed limits displayed, during rain or fog, on six VMSs on the A31 Motorway. This is a part of the Melyssa project (funded by the European DRIVE II program). The French Motorway company "SAPRR" has managed the equipment and the trials and the French "CETE de l'EST" has proposed the speeds limits values in function of the rain or fog levels. Some results are given (only for rain periods) on the speed reduction and on the lane assignment when the VMSs are switched on; lane assignment strategies differ within the drivers, and seem to be sensitive to the rain, to the speed observed on the slow lane, to the interference of trucks during the rain. (Author/publisher)
Abstract