Research On Anti-collision Technology With VANET And Multi-sensors Fusion

With the increasing of human life quality,vehicles become one of the most commonly used means of transportation,and the traffic safety is becoming more and more serious.The research on the anti-collision safety technology of the vehicle has attracted more and more attention from the major automobile manufacturers and many scholars.At the same time,the concept of intelligent transportation system(ITS)has gradually entered the area of road traffic construction,and more and more countries begin to layout the future traffic system based on ITS.As one of the important technologies of ITS,the vehicle ad-hoc network(VANET)technology has developed rapidly,and causes a new research field on the anticollision safety technology.Based on above reasons,this paper regards vehicle distance measuring technology as the starting point and studies the anti-collision technology with VANET and multi-sensors fusion.Firstly,the research background and related basic technologies development status were introduced,which provided the basis for the research of anti-collision technology with VANET and multi-sensors fusion.The related basic technologies include VANET,multi sensor information fusion technology,VANET structure,mobility model and sensor system.Secondly,this paper analyzed the clustering algorithm of VANET,and proposed a locationbased VANET clustering algorithm and a fast information delivery strategy.The algorithm is a kind of information sharing scheme based on VANET in the highway environment,it realizes the function of the nodes in the cluster to share their own information to the rear nodes.By adopting the unidirectional transmission for shared information,and dynamically changing the clustering distance threshold with the variance of vehicle density,the proposed algorithm can effectively avoid two kinds of problems caused by improper selection of clustering distance threshold.One is that the cluster cannot effectively cover the vehicles among which the distances are less than the safe distance,and the cluster is too small to achieve the purpose of sharing the message effectively.The other is that too many vehicles are within one cluster,which causes too many information for transmission,too large message forwarding delay,etc..The feasibility and correctness of the algorithm were verified by analysis and simulation.Finally,based on the research and analysis of the target moving model,the measurement model,the theory and algorithm of Kalman filter,we propose a method to indirectly measure the target vehicle with the information shared by the front inner nodes,which made the front external nodes to be seen earlier and the measurement of the distances more accurate.The proposed method combines the proposed clustering algorithm and the existing information fusion algorithms,and can be divided into two steps,the first-fusion and the second-fusion.The first-fusion can achieve a higher accurate distance results and the second-fusion eliminates the error data and fusions the multiple first-fusion results.Simulation results are provided in this paper to verify the effectiveness,feasibility and correctness of the proposed method.