Recognition Of Dangerous Driving Behavior Based On Lane-level Positioning

In China,the serious traffic accidents coursed by the vehicle's driving occur frequently.Studies show that the Illegal lane changing,retrograde,driving in the wrong lane and some other dangerous driving behaviors mainly result in the traffic accidents.For a long time,that the fewer applications of the IT in traffic safety management,the lack of the recognitions and the supervisions of the driving behaviors,the lack of efficient data collecting method of the driving behaviors result in the lack of the pertinence and the validity of the traffic safety management.Related research work about dangerous driving behaviors are carried out currently at home and abroad,most of these studies were performed using image recognition technology or acquiring the information such as the driver's physiological characteristics to achieve the dangerous driving behaviors identification.Series of preparation need to be done in this technology before conducting the research like digital video recorder and physiological characteristics acquisition sensors mounting in the automobile.These increase the research cost and some obstacles still exist restricting the application of this technology.Meanwhile,Satellite navigation technology is developing rapidly and the research based on the satellite positioning data in this filed has been found also in deficiency.Therefore,this paper conduct the dangerous driving behaviors identification research combining with the satellite positioning technology in the lane level.The main work has the following points:(1)Algorithm design for dangerous driving behavior.:In this paper,based on the lane level positioning and the characteristics of high precision map,analysis of the behavior characteristics about the risk of the lane changing,illegal riding lines,retrograde,and not in accordance with the provisions of the lane.According to the several kinds of dangerous driving behavior's characteristic to design the recognition algorithm in a lane-level positioning.(2)Development of dangerous driving behavior recognition software.:In this paper,based on the android operating system,in combination with electronic maps,satellite navigation,develop a dangerous lane changing,specifically for pressure,retrograde,navigation lane by the regulation of dangerous driving behavior recognition system.The system can not only identify dangerous driving behavior,also to make a statistics,the number of dangerous driving behavior for the analysis of dangerous driving behavior data provides a convenient way of gathering.(3)Based on dangerous driving behavior analysis of vehicle traffic threat level assessment: In this paper,based on the developed software of dangerous driving behavior recognition,on the basis of the type of vehicle risk index system,the use of dangerous driving behavior recognition software statistics of the number of dangerous driving behavior as the research object.Using the method of AHP and matter-element completed for vehicle traffic threat level assessment.Through the case analysis,the model can evaluate the risk of to a vehicle.By on line identification of dangerous driving behavior,it can warn the driver in time and improve the safety of the driving process.At the same time,through the system of driver's dangerous driving behavior for statistical analysis,to formulate strategy of safety management of dangerous driving behavior,driver's driving behavior norms,reduce the occurrence of traffic accidents.