Desenvolvimento de aplicativo de dispositivo móvel para ajustamento de coordenadas geográficas

Abstract

Mobile devices, as smartphones and tablets, are widespread worldwide in society reaching a wide range of users due to its multifunctionality and constant technological evolution. One of the main applications of those devices is the geolocation by GNSS receiver in mobile devices. Although the technological evolution of GNSS receivers in mobile devices has been exponential, one of its main limitations is the positional accuracy of these receivers. The average of positional accuracy of smartphones and tablets is estimated between 3 to 10 meters, depending on environmental factors and their hardware and software. Thus, this work aims to improve the static positioning accuracy by using statistical properties of GNSS observations, applying the technique of geodetic observations adjustment in an app developed for mobile devices. The app was developed on the Android platform with the capacity to store information about geodetic marks, to collect geographic coordinates and apply the technique of geodetic observations adjustment by the parametric method. To test the app, was realized point surveys by static positioning over geodetic marks, using this mark as a comparing reference, and the last nine marks were used how adjustment parameters. On the surveys was obtained 2000 geographic coordinates over each point using about 40 to 50 minutes for each point survey. As results on app tests, improvements on positioning accuracy were acquired after applying the adjustment on the geographic coordinates observed, getting sub-metric accuracy in two ideal cases with 66,6 and 92,6 centimeters as the positional difference from the reference mark. Furthermore, the other four points shoed accuracy better than 1,78 meters. Nevertheless, the last four points showed positional accuracy worse than 2,2 meters, reaching out 5,39 meters. However, since the spatial distribution of the observations, was percept spatial deviations from exceeded systematic errors, mainly from multipath, due to the existence of edification and vegetation in the neighborhood of the marks. It should be highlighted that the accuracy of observations was impaired, but the precision was good because the observations were just displaced from the reference due to systematic errors. In the distribution that the existence of the errors was low and the precision was good, the adjustment improved considerably the positional accuracy of the point. On the other hand, the adjustment improved over 49% the positional accuracy in all points when compare the adjusted coordinates to mean error of the observation (RMSE), reaching out 81%. Thus, the adjustment was improved considerably the accuracy of the positioning, reaching out to sub-metric quality when the systematic error was not harmed the statistical and spatial distribution of observations errors. Besides that, the adjustment showed positional quality better than the observations RMSE, offering a more accurate positioning than the average of the set of observations of each point.