The Relative Position Of The Vehicle-mounted Network Design And Implementation Of A Prototype Vehicle Location System

Traffic accident problem has become the focus of social problems. Localizationtechnology is one of the most important applications. The traditional localizationtechnologes cannot be used in many occasions, for example, bad weather (snow, rainetc.), bad light conditions (dark, sun glare etc.) as well as in presence of urban noise (e.g.faded lane marks, parked cars and crash barriers etc.).Other localization technologyhave high equipment cost, can only locate neighbor vehicles, not suitable for installationin the existing cars. Aiming at these shortcomings, this paper puts forward a novelvehicle localization technology using multiple wireless sensor nodes, presents a vehiclerelative localization system framework, and within the framework, studies the main keytechnologies of the system.In order to reduce the effect of the single RSSI model and the average model tofinally RSSI value. This paper uses gauss model to remove part of the greatest error ofRSSI value and uses optimal weighting method to obtain the final RSSI values. Theexperimental results show that, the error of the gauss weighted model is smaller thanaverage localization model.Through the verification experiment, it is concluded thatwhen two vehicles are in the same lane the link loss between sensor nodes is symmetric,however, when two vehicles are in the different lane the link loss is asymmetric. Wedesigned two kinds of relative localization algorithm of sensor location model, as wellas the relative localization optimization algorithm. Through the experiment, the bestposition of the sensors on the vehicle body is determined. The response time of thesystem can’t more than7s. Under different environmental conditions and differentvehicle types, the average localization accuracy can reach90%. This paper presents aranging algorithm based on RSSI and distance mapping. Change the traditionalone-to-one relationship between single RSSI values and distance to one-to-many relationship. Using one-to-many mapping relationship, calculate the distance. Thisranging methodwithin50meters, more than90%can achieve within2meters inerror,and the average relative error is3.99%. The localization system can locate understraight and curve road condition. It can locate neighbor vehicles and non-neighborvehicles. Sensor equipment is easy to install on the existing vehicles withoutconsidering the condition of light and cloudy.