Analysis Of Motor Vehicle Positioning Algorithm Based On Non-linear Statistical Model

How to achieve accurately and rapidly vehicle positioning is an important issue which modernintelligent transportation systems must to be researched. Usually, the vehicle positioning can bedivided into two stages. The first stage is the preliminary positioning stage, which get the vehicleposition information through the satellite or wireless network. The second stage is the filtering stage,which using an appropriate filtering algorithm to filtering the position information. The purpose ofthis paper is to research the appropriate filtering algorithm. Then uses the algorithm to filtering thepreliminary positioning information, by this method the accuracy of vehicle position can beimproved.In order to improve the accuracy of vehicle positioning, the information of vehicle away isused as a part of observations to establish non-linear statistics vehicle motion model. UnscentedKalman Filtering algorithm can adapted to nonlinear systems well, at the same time it also has bothhighly positioning accuracy and lower computational complexity, so Unscented Kalman Filteringalgorithm is chosen as the based filtering algorithm to improve the accuracy of vehicle positioning. Interacting Multiple Model algorithm is an adaptive algorithm, it can be described the states of thesystem well when the system states changed. This function the Unscented Kalman Filteringalgorithm does not have. So, in order to solve the problem of error increased when vehicle motionstates changed, the Interacting Multiple Model algorithm is used to combining with the UnscentedKalman Filtering algorithm. Through the experiment, it can be seen that the positioning accuracy ofInteracting Multiple Model Unscented Kalman Filtering algorithm is obviously better thanUnscented Kalman Filtering algorithm, within the scope of this page discussed, the improvedeffects can be achieved about twenty percent.As an improved algorithm, Interacting Multiple Model Unscented Kalman Filtering algorithmprovided a way to improve the accuracy of vehicle positioning.