Study Of Anti-collision Warning System Based On Wireless Communication

With the rapid growth of running vehicles in modern society,accidents caused by automobiles are rising fast and the number of human injuries and money loss reach a new high level.Although modern safety technology can guarantee driver and pedestrian’s life,for instance,passive safety technologies like safety belt and airbag could minimize the injury level when accidents happened,active safety technologies like ABS,ESP,AEB could maintain vehicle’s driving stability as well as minimizing the threat towards pedestrians and surrounding vehicles,yet those active safety technologies depend on expensive sensor,which is not applicable to all vehicles.In recent years,as communication technology keeps evolving and maturing,connected intelligent vehicle and intelligent transportation system are becoming the new buzzword,countless research institutes and college scholars started studying wireless communication and establishing dedicated experimental area,using wireless communication to improve vehicle safety level is becoming the new researching focus.Vehicle anti-collision warning system means on-board warning system can not only inform driver of surrounding driving situation,like vehicle location and driving speed,but also detect potential accidents and give driver alerts.Wireless communication indicates different devices could establish connection,achieve data exchange and information sharing without physical link,like cellphone communication.Combining these two technologies,vehicle can “talk” and share information with each other after connection,and sensing range is extended to improve vehicle driving safety.This paper considers wireless communication and anti-collision warning system as core issues,focusing on the study of vehicle to vehicle communication,vehicles relative positioning and driving safety calculation.This paper first discussed current environment perceptive technologies and analyzed advantages and disadvantages,proposing obtaining useful information via wireless communication to fulfill pre-warning,then introduced domestic and international studies about connected intelligent vehicle and anti-collision warning system in recent years,and started the research in this paper based on those predecessors.Chapter two mainly contained hardware structure and software requirements of on-board warning system,to be specific,including ZigBee technology that achieves vehicle communication,GPS technology that achieves relative positioning,human machine interface that gives driver alerts and communication protocol that ensures correct data transfer among different devices.Based on selected devices and designed protocol,a vehicular ad-hoc network platform was proposed.System modeling was elaborated in chapter three,including ZigBee communication model that fulfills vehicle communication,improved safe distance model that constantly evaluates longitudinal driving situation,GPS positioning model that locates surrounding vehicles,and driver intention recognition model,therein,communication model,safe distance model and positioning model are core ones.Communication model is mainly responsible for ensuring correct data transmission among different devices,including data transformation between CAN bus and serial port,relevant data acquisition from CAN bus.After communication model,modified safe distance model was proposed since brake-process-analysis-based traditional safe distance model has a few problems,and warning strategies were also briefly introduced.Vehicle positioning model mainly focused on GPS coordinate transformation and unscented Kalman filter algorithm was also used in this paper to improve positioning accuracy.Simulations and outcome analysis of each model,main system functions’ outdoor verification under campus environment were presented in chapter four.Communication tests,including static and dynamic link tests between nodes,were used to evaluate accuracy and stability of ZigBee module.In order to validate safe distance model’s reliability,outdoor tests were conducted when driving speed was below 40km/h,otherwise,simulation tests were conducted to verify the model.GPS model validations also contained static(vehicle not moving)and dynamic(vehicle moving)tests.Like safe distance model,driver intention recognition model consisted of outdoor tests(vehicle speed below 40km/h)and simulations(vehicle speed over 40km/h).Four typical driving situations were tested outdoors,including acceleration,deceleration,lane change and turn.Summarization and future works were presented in the last chapter.