Decentralized Cooperative Lane-Changing Decision-Making for Connected Autonomous Vehicles

In this paper, we proposed a decentralized cooperative lane-changing decision-making framework for connected autonomous vehicles, which is composed of three modules, i.e., state prediction, candidate decision generation, and coordination. In other words, each connected autonomous vehicle makes cooperative lane-changing decision independently. In the state prediction module, we employed existing cooperative car-following models to predict the vehicles' future state. In the candidate decision generation module, we proposed incentive based model to generate a candidate decision. In the candidate decision coordination module, we proposed an algorithm to avoid candidate lane-changing decision that may lead to a vehicle collision or traffic deterioration to be final decision. Moreover, the effects of decentralized cooperative lane-changing decision-making framework on traffic stability, efficiency, homogeneity, and safety are investigated in a numerical simulation experiment. Some stability, efficiency, homogeneity, and safety indicators are evaluated and show the high potential of our proposed framework in traffic dynamics.