A Decentralized Multi-Criteria Optimization Algorithm for Multi-Unmanned Ground Vehicles (MUGVs) Navigation at Signal-Free Intersection
Résumé
In intelligent traffic management, one of its core challenges lies in Multiple Unmanned Ground Vehicle (MUGVs) navigation in cluttered and dynamic environments. To be viable, a multi criteria navigation scheme is required to deal with several critical situations. With its relative low execution time, the Probability Collectives (PC) algorithm has succeeded in generating fast and feasible solutions when applied to manage challenging scenarios, such as in signal -free intersections and roundabout (see Philippe et al. (2019)). Indeed, PC is an interesting decentralized approach for general cases that enables to manage complex systems under probabilistic hypotheses. However, the PC is sensitive to uncertainty in the navigation process, which highlights the need to adopt several safety margins. These margins permit vehicles to adapt their dynamics and to react properly to unexpected events. Accordingly, the present work aims to integrate a reliable risk management strategy into the PC algorithm by introducing a novel &constraint PC method. With the enhancements integrated into an already existing approach, the &PC based navigation strategy is able to obtain a better fusion strategy with considering both efficiency and safety. Accordingly, this work aims to develop an appropriate balance develop a proper balance between the high-quality solution and acceptable computational speed. Further, typical scenarios of unsignalized intersection have been used intensively in simulation setups to demonstrate the efficiency of the proposed approach