Lane-Changing Decision-Making Model for Automated Vehicles Based on A Novel Lane Attractiveness Index

Reasonable lane-changing decision-making strategy is an important part of developing automated vehicles. Aiming at the problem that traditional lane-changing decision-making only considers the driving status of adjacent vehicles, this study proposes a lane-changing decision-making model based on the lane attractiveness index. The lane-level attractiveness index is constructed based on the information on adjacent vehicles and a larger range of surrounding vehicles, and the constraints of lane-changing feasibility factors and risk factors are added to ensure the safety of lane-changing decisions. The simulation results show that the proposed lane-changing decision-making model is consistent with the actual behavioral decisions extracted in the highD dataset and is validated in a basic driving scenario.