Research On Pedestrian And Vehicle Collision Warning Method Based On Edge Computing

At present,there are many private cars in China,and the corresponding road traffic flow is increasing.However,the overall quality of motor vehicle drivers has declined,and the safety awareness of road users is still insufficient,resulting in increasing safety accidents year by year.Among them,pedestrians are susceptible to injuries in pedestrian accidents because pedestrians are not protected by special devices.As a result,the number of pedestrian deaths has increased year by year in road safety accidents.In view of this situation,in order to ensure the safety of pedestrians,the pedestrians and vehicle drivers are warned by studying the communication between pedestrian equipment and vehicles to reduce the traffic accidents caused by the lack of pedestrian vision and the distraction of pedestrians by mobile phones.This paper uses an edge computing server to realize the vehicle's perception of the surrounding environment,and proposes a vehicle pedestrian collision avoidance system(MECA)based on edge computing.The main work of this article is as follows:1.MECA uses the mobile edge server as an intermediate server to collect real-time data information of pedestrian devices and vehicles within its range,indirectly realizing the communication between vehicles and pedestrian devices(UE),so that pedestrian devices also participate in road safety applications.as a center.Pedestrian status information is obtained through pedestrian devices,and vehicle information is obtained through vehicle-mounted units.2.After the mobile edge server collects the information,the extended Kalman filter is used to accurately predict the future positions of vehicles and pedestrians respectively,rather than fuzzily demarcating the danger zone,which makes the division of collision points more accurate and makes collision warning more accurate.Kalman filtering is the best algorithm to estimate the state of the system from uncertain and indirect measurements.The extended Kalman filter developed from this is widely used to fuse the data of various sensors to estimate the position of nonlinear moving objects.Practice has proved that the extended Kalman filter sensor can get the best estimate of the precise position of the moving object and can effectively eliminate the error generated by the sensor.3.After predicting the positions of pedestrians and vehicles,if a collision occurs,a collision time range(TTCR)is established as a warning indicator based on the distance between pedestrians and vehicles to the point of possible collision and the current speed to resolve the effects of system errors and vehicle length.Simulation results show that the proposed trajectory prediction algorithm can provide accurate motion relationships between vehicles and pedestrians.At the same time,the proposed collision avoidance algorithm can effectively reduce the collision risk.