Integrated Motion Planning for On-Ramp Merging Based on Stackelberg Game Modeling Considering Interactive Characteristics

Efficient and safe on-ramp merging is crucial for mitigating traffic congestion and enhancing vehicle safety. In this paper, an integrated framework for decision-making and trajectory planning in on-ramp merging scenarios is proposed. First, adequate longitudinal spacing can be established by adjusting vehicle velocity prior to merging if the initial safety space is insufficient. Then the combined framework of the Stackelberg Game theory and sampling-based planning method is developed to enable simultaneous decision-making and trajectory planning for on-ramp merging. Specifically, the potential effect of different merging behaviors of the ahead vehicle on the predicted motion sequence of the subject vehicle is comprehensively considered. Moreover, driver aggressiveness is accurately modeled by the proposed hierarchical identification method that integrates offline LSTM neural network training with online modification based on utility maximization reasoning. Finally, the effectiveness of the proposed algorithm is verified under various simulation scenarios and human-in-the-loop experiments.