Jointly conducted by UGB and IRD, this research proposes a multi-agent model to optimize urban mobility in Marrakech.
Saâd Touhbi, researcher at UMMISCO (Université Gaston Berger de Saint-Louis Sénégal).
Thesis defended in 2018 at the doctoral school École doctorale Informatique, télécommunications et électronique de Paris (1992-....).
This research is the result of an international co-supervision between several partner institutions.
Referenced in the ABES/STAR network, this thesis meets the rigorous criteria of French higher education.
Urban mobility is a major challenge for developing cities, and Marrakech is no exception. Indeed, with a rapidly growing population that has reached nearly one million inhabitants, and often saturated infrastructure, traffic management becomes crucial to ensure the smooth functioning of the city. Traffic jams, delays, and pollution are daily issues that affect not only the quality of life of citizens but also the local economy. Saâd Touhbi's thesis, developed in co-supervision between Gaston Berger University of Saint-Louis and IRD, proposes a multi-agent model for synthetic traffic generation. This innovative model allows for the simulation of driving behaviors and anticipates vehicle flows on the roads of Marrakech, thus offering a potential solution to the challenges of urban mobility.
The methodology adopted by the author is based on a rigorous four-step approach. First, the processing of vehicle arrival data and the calculation of the time interval between vehicles (TIV) are carried out. This step is crucial as it allows for understanding how vehicles interact on the roads, particularly during peak hours. Next, samples are defined to represent different traffic levels, ranging from low traffic periods to peak hours when traffic is particularly dense. The third step involves selecting and estimating probabilistic models, allowing for the prediction of future behaviors based on historical data. Finally, a comparison is made between the estimated models and empirical data, thus ensuring the validity and reliability of the developed tool. This methodological approach allows for the validation of the traffic generation tool, which proves capable of producing traffic levels consistent with the desired configurations.
The results of this research are particularly revealing. They show that the Pareto IV model is suitable for all traffic levels, while other models, such as the Exponential, are found to be inadequate. These conclusions pave the way for practical applications, particularly in simulating air quality based on traffic, which is essential for public health and environmental policies. Indeed, the developed tool allows for estimating automotive pollutants without requiring point data, representing a time and resource saving. For example, decision-makers can now anticipate pollution peaks and implement preventive measures, such as traffic restrictions or awareness campaigns.
It is imperative that decision-makers take these advancements into account to improve urban traffic management. Implementing this model could not only reduce traffic jams but also contribute to reducing pollution. By integrating these tools into public policies, governments can better anticipate infrastructure needs, such as the construction of new roads or the improvement of public transport, thus enhancing the quality of life for their citizens. Furthermore, this research could serve as a model for other cities in West Africa, where urban mobility challenges are similar. Cities like Dakar, Abidjan, or Ouagadougou, which also experience rapid urbanization and traffic issues, could benefit from the application of this multi-agent model.
In summary, Saâd Touhbi's thesis represents a significant advancement in the field of urban traffic modeling. By offering a rigorous and adaptable method, it opens promising perspectives for managing mobility in rapidly changing urban contexts. The implications of this research extend beyond Marrakech and could be applied to other cities in West Africa, thus contributing to a better understanding and management of urban traffic dynamics. Ultimately, the integration of these tools into public policies could transform the way cities address mobility challenges, promoting sustainable development and improving the quality of life for all.
Access the full study
Sources and Access
Saâd Touhbi. Élaboration d’un modèle multi-agents pour la génération synthétique de trafic : application à la mobilité urbaine de la ville de Marrakech. Intelligence artificielle [cs.AI]. Sorbonne Université; Université Cadi Ayyad (Marrakech, Maroc), 2018. Français. ⟨NNT : 2018SORUS326⟩. ⟨tel-02864774⟩