ABSTRACT
Road traffic is a complex phenomenon and its collection follows a spatio-temporal process. Its management has attracted the attention of academia and industry because it often suffers from data gaps caused by the temporary deployment of sensors, malfunctioning detectors, and poor communication systems. In Morocco, road traffic data management is managed by the “National Center for Road Studies”, which is attached to the Ministry of Equipment and Water. The center collects a large amount of data on the traffic flow on the main roads in Morocco. To obtain the traffic volume AADT (Annual Average Daily Traffic), which is the common traffic indicator, it is necessary to obtain continuous data. However, it is not possible to collect data on all roads because of the cost involved. Therefore, the center uses periodic counters with a counting time of 8 days per semester (16 days per year). These data, together with those obtained from the permanent counter, are then used to calculate the estimated annual traffic characteristics and seasonal variations. In this context, and to fully exploit the history of road traffic information and its spatio-temporal correlation, this chapter presents an approach based on artificial intelligence methods to estimate the missing values related to road flow in Morocco. The simulation results show that the adopted approach allows for estimating the missing data with high accuracy. The present work will contribute to improving the quality of the road traffic collection because the missing data are likely to have an important weight and therefore cannot be ignored during the statistical analysis of road traffic.
