Research On GPS Algorithm Based On Outliers Filter For Vehicle Mileage

In the early twenty-first century,The Internet of things technology is coming into people’s sight slowly.With the gradual maturity of the development of the Internet of things,Vehicle networking technology has also been pushed on strongly.The current vehicle networking market is mainly composed of the front-mounted vehicle networking market,which has cloud services platform for foreign investment enterprises to support and design dominantly,the vehicle mileage statistics has been one of the design which can’t be ignored in vehicle monitoring system,especially in the assessment of vehicle quality,assessment of driver’s behavior,the vehicle mileage is very important reference index.In this thesis,concentrate on the vehicle real-time GPS sample data for study,causing GPS sample data outliers filtering problem.According to the process of GPS sample data filtering,this thesis puts forward a kind of technical framework of vehicle mileage statistics using GPS data.The main contents are listed as follows:(1)In this thesis discusses the map matching technology which combines the electronic map and the GPS sample data.(2)In this thesis,an extended algorithm based on circular sieve filter algorithm is proposed,and the Calman filter is used to filter the outliers of GPS sample data.(3)Several vehicle mileage statistics methods are studied to find out which one is the most suitable for this system.(4)Design and accomplish the sub module software development of the "real time map monitoring",and analyze and test the results of vehicle mileage.After discussing and testing of the above process.The software implementation of the real time map monitoring module is completed.The RabbitMQ data reception mechanism,GPS sample data outliers filtering algorithm is integrated into the whole design process,At the same time,the map matching technique is used to combine the GPS data and the digital map,and in this thesis,the statistical accuracy of the cumulative polyline method is expected to be the final result.