A lane level positioning-based cooperative vehicle conflict resolution algorithm for unsignalized intersection collisions

• A lane level positioning solution for vehicle situation perception is proposed.
• E-Map-based map matching increases cost efficiency of on-board system.
• A location-based cooperative vehicle conflict resolution mechanism is proposed.
• The mechanism relies on lane level positioning and PSO-based control strategies.
• The emergent control strategies consider both safety and driver behaviors.

We consider the problem of vehicle operation safety at unsignalized intersections. The scheme of cooperative vehicle infrastructure system provides a promising solution to safety-related traffic issues. In this paper, we propose a novel cooperative vehicle conflict resolution algorithm for unsignalized intersection collisions. An enhanced road map is utilized to achieve lane level position precision. Because of the integration of map data, few sensors are required, which promotes a simple and cost-efficient solution for perceiving vehicle situations. Furthermore, we introduce an evaluation method for describing the emergency degree of ongoing collisions. Particle swarm optimization is employed to calculate the target acceleration and control the vehicle to prevent unexpected collisions. The results from experiments and simulations indicate that the proposed algorithm achieves better perception compared to similar previous strategies and illustrates the effectiveness of maintaining a vehicle in safe motion using emergent braking control. The simulation results also indicate the algorithm’s tremendous power for driver assistance in practical unsignalized intersection environments.

Figure optionsDownload as PowerPoint slide