Smooth lane-changing: a game-theoretic approach incorporating pre-game intentions

Combining lane-changing intentions with a game-theoretic approach is a novel outlook on vehicle dynamics and decision making in traffic. Because identification also considers both speed and distance, the detailed analysis of lane-changing behavior is possible. Moving away from conventional car-following requisites, the model employs the Sigmoid Intelligent Driver Model (SIDM) and forces a fifth-degree polynomial for quality lateral acceleration transition during lane changes to allow variations in styles of driving. The decision-making process is completed by a generalized framework for the development of more meaningful and interpretable utility functions through monotonic and dimensionless transformations of quantitative variables. Random perturbations are integrated into the payoff matrix to address the inherent uncertainty in real-world driving environments, ensuring the robustness and practicality of the model under various conditions. The validation conducted under two distinct traffic scenarios confirmed the effectiveness and elasticity of the model, indicating potential improvements in traffic conditions.