Mutual Exclusion Algorithm for Real-time Traffic Dispatching Based on Game Theory

Abstract

With the rapid increase of vehicles, real-time dynamic dispatch is an important means to alleviate traffic congestion. Game theory-based real-time vehicle scheduling algorithms are often used for dynamic adjustments, but traditional algorithms only switch signal phases for fixed phases during dynamic adjustment scheduling, and cannot achieve the comprehensiveness of real-time scheduling; or phases with large differences in traffic flow when setting the duration, the reasonable waiting time for the smaller traffic phase are not considered. In order to solve the real-time dynamic adjustment of signal lights, an improved algorithm based on game theory is designed to control the signal switch in the game mode, and the waiting time of vehicles at traffic intersections can be reduced by adding parameter values of lane waiting time changes. Through experimental comparison, the algorithm reduces the vehicle stay by 3.4% and the waiting time by 4.4% compared with the fuzzy algorithm of the polling mechanism.