Study On Cooperation Algorithm For Multi-intelligent Vehicles

Multi-Intelligent vehicles cooperation is an important aspect of intelligent vehicle technology, the vehicles cooperation can significantly improve the system's safety and efficiency. This paper studies on the design and testing issues of the intelligent vehicles cooperative algorithm, and also involves design and implementation of multi-vehicles simulation platform.Platoon cooperation is a focus of the multi-Intelligent vehicles study, this paper puts forward a decentralized control method for platoon based on the spring-damper systems. The method modeled the platoon as a group of vehicles connected with spring damper systems, thus vehicles use only local bi-directional information; the method also decoupled the system using a special overlapping structure decomposition method; it used decentralized LQ control method, shrank the expanded system back to the original system, to find the spring damper parameters for the sub-optimal solution. Simulation compared this approach with the traditional control methods to prove its effectiveness and good results.Intersection coordination is another important aspect of multi-vehicle cooperation, the roundabouts and the crossroads are and the two typical scenarios. For the intersection, this paper improved the conflict table algorithm without traffic lights, and designed the experimental interactive system for algorithm comparison; the simulation platform was used for the comparison of this algorithm and the algorithm with traffic lights. For the roundabout, this paper presents advance planning method based on MSRS with out stop; a large number of simulation results show that the algorithm has energy-saving effect. The actual experiments for multi-Intelligent vehicles cooperation have strict requirements, technical complexity, and they are not safe. Thus the simulation software platform is needed. This paper designs and implements CyberTORCS simulation platform for platoons, and improves the intersection interactive system of the multi-vehicles; for the roundabouts the paper designs simulation environment for multi-intelligent vehicles algorithm based on MSRS which is an appropriate software environment for research.