A game theoretic framework using Nash strategies for urban traffic control and path planning algorithms for vehicles have been proposed in the paper. The simulation results underlie the intuition that a game theoretic strategy is able to outperform constant green times strategies. The solution is computationally expensive and can be realized only if some simplifications are carried out. The most important simplifications appear in bundling junctions into groups and decreasing the number of decisions. The game theoretic framework is able to incorporate path planning algorithms for vehicles. Two algorithms have been discussed. A deterministic path planning provides the optimal path between two road links in the traffic network while a probabilistic approach is able to achieve better performance when the path planning method is carried out for many vehicles simultaneously.
Urban traffic control and path planning for vehicles in game theoretic framework
Stadtverkehrlenkung und Wegplanung für Fahrzeuge in einem spieltheoretischen Rahmen
Robot Motion and Control 2007, RoMoCo, International Workshop on Robot Motion and Control, 6 ; 437-444
2007
8 Seiten, 3 Bilder, 10 Quellen
Conference paper
English