Research And Implementation Of Urban Traffic Jam Warning Based On Vehicle Ad-hoc Network

With the all-round development of our country’s economy,the number of cars has increased year by year,and the relevant requirements for vehicles in communication are getting higher and higher.At present,the frequent occurrence of car accidents has not only brought unnecessary traffic jams,but also caused many inconveniences to people’s lives,economic losses,seriously threatened life safety,and so on.In order to solve the above problems,this paper focuses on the image recognition in the vehicle collision scene and the communication of the vehicle after the collision.The main work of this paper is as follows:(1)Introducing the communication channel model and the algorithm of image recognition when the vehicle is driving.First,the classification of mobile communication is introduced,and the mobile characteristics of vehicle communication is explained.Secondly,it introduces some fading channels including Rician fading channel and Nakagami-m fading channel,analyzes and explains the Probability Density Function(PDF)of these two channels.Then it introduces the two mobility models in mobile communication:Random Waypoint Mobility(RWP)and Random Direction Mobility(RD),and clarified the node movement paths and PDF of these two mobility models.At the same time,it introduces the working method of Maximal Ratio Combining(MRC)technology in diversity technology,and the outage probability and the Bit Error Rate(BER)are analyzed.Finally,when collisions are happened,the YOLOV3 algorithm is used to collect field photos for collision recognition.(2)Analyzing the RWP mobile model under Rician fading channel,and giving the accurate expressions of outage probability and BER.At first,it presents the scene when the vehicle uses the RWP mobile model to communicate.Then,according to the PDF of the Rician fading channel and the PDF of the RWP mobile model,a PDF and Cumulative Distribution Function(CDF)of the Rician fading channel in a mobile environment are obtained.On this basis,the expressions of outage probability and BER are calculated.The influence of different dimensions on outage probability is also considered.Compared with η-μ fading channel,the superiority of Rician fading is verified.(3)Analyzing the outage probability under Nakagami-m fading channel.Firstly,tin order to overcome the influence of multipath fading,the effect of MRC diversity technology on the Signal-to-Noise Ratio(SNR)of the output signal is considered,and a calculation method to improve the SNR of the channel is proposed.Then this paper calculates the PDF of the RD mobile model under the Nakagami-m fading channel with MRC diversity technology.Secondly,this paper calculates the formula of the outage probability,and the influence of the RD mobile model on the outage probability in different dimensions is also considered.Through simulation analysis,the correctness of the proposed expression is proved.(4)Analyzing the anti-blocking warning scheme based on neural network after a vehicle collision.Using the incident camera that comes with the passing vehicle,it automatically takes pictures of the car accident scene and recognizes it.The problem of the accuracy of image recognition at the scene of the collision of the vehicle is studied.The vehicle collision scene data set is established,and the data set is put into YOLOV3 for training to obtain the image recognition accuracy.In order to improve the YOLOV3 algorithm,this paper introduces the SENet module into YOLOV3 to enhance its image recognition accuracy.Through simulation analysis,it is concluded that the improved algorithm has a higher image recognition rate,which can accurately identify the images of vehicle collision,and can achieve the goal of anti-blocking warning.