Collision Avoidance System For Unmanned Vehicles Considering The Comnuncation With The Vehicle In An Entrance Ramp

One of the major significance of the development of unmanned vehicle technology is to improve the security of transport systems. Entrance ramp area has high incidence of traffic accidents, so the study of the entrance ramp collision avoidance system for unmanned vehicles has an important significance for the development of unmanned vehicle technology. In this thesis, the collision avoidance system for unmanned vehicles was carried out under the entrance ramp environment.This thesis figured out the key advantages/shortcomings of prior solutions and the potential risks of collision based on the driving rules in the entrance ramp. Then a novel method combining information of detection sensor and V2 V communication was proposed. Co-simulation experiment using Pre Scan and Simulink is conducted. The experimental results show that our method can lead unmanned vehicle to pass the entrance ramp safety in the case of social vehicle perform a clear driving intention.V2V communication was used to share information between host vehicle and social vehicles, multi-dimension hidden Markov model(MDHMM) whose parameters were trained beforehand was used to estimate social vehicles driving intention. Hybrid state system(HSS) was used to control the intelligence vehicle while considering its decision making, states and dynamics. Co-simulation using Pre Scan and Simulink is conducted. The experiment results show that the intelligent vehicle behaved a commendable social behavior performance in the entrance ramp.The hardware and software program of collision avoidance system were designed based on the platform of Dong Feng Motors in order to verify the collision avoidance system under the real condition(the real unmanned vehicle in the real circumstance). The experiment platform for research combines two vehicles--the unmanned vehicle and a manned vehicle. GPS and Inertial Navigation System(INS) were used to get the vehicles’ position and heading information in the platform, Zig Bee modules were used to communicate between the vehicles, and the ESR millimeter-wave radar was used to detect the obstacle vehicles. The experiments results showed that the collision avoidance system base on V2 V and radar was able to be used in the real entrance ramp.