Cooperative lane-changing decision-making and optimization with uncertainties in the vicinity of signalized intersections for mixed traffic: a cyber-physical system perspective

To address issues such as traffic flow instability caused by lane-changing behavior in the vicinity of intersections under mixed traffic, this paper proposes a cooperative lane-changing decision and optimization method for intelligent vehicles (IVs) from a Cyber-Physical System (CPS) perspective. Considering uncertainties, a bi-level planning model integrating macroscopic optimization and microscopic response is constructed within the CPS framework. In the cyber layer, the upper-level model minimizes the total travel time near the signalized intersection, while the lower-level model minimizes individual predicted travel time subject to upper-level guidance and safety constraints. The physical layer executes the resulting lane-changing decisions and feeds the execution outcomes back to the cyber layer, forming a closed-loop macro–micro optimization mechanism. Finally, simulations were conducted using a real-world signalized intersection. Results demonstrate that the proposed method significantly improves operational efficiency while maintaining favorable driving safety performance.