The 20th International Technical Meeting of the Satellite Division of The Institute of Navigation

drahmadtaha's picture
Research Title: 
A Continuous Updating Techniques for Loosely Coupled RTK GPS with Total Station Observations
Authors: 
Taha A.
Country: 
USA
Date: 
Mon, 2007-09-24
Research Abstract: 

In urban areas, satellite visibility can limit the exclusive use of GPS to complete a survey. To overcome this, Real Time Kinematic (RTK) GPS could be used to establish control points while total stations are used to complete the surveying. Combining GPS with a total station in one single instrument (i.e. Leica SmartStation) reduces the surveying time, the cost and increases the overall accuracy. All set-up possibilities within the SmartStation, described by [4], requre having a minimum of two inter-visible control points. However, when positioning in densely settled urban areas, it is difficult to find two inter-visible RTK fixed points with good geometry. Furthermore, errors could be accumulative if a technique of traversing is used to propagate coordinates from one point to another. In contrast, it is possible to have several points where RTK positions are available. This paper presents the developed Continuous Updating Technique (CUPT) to loosely coupled GPS RTK positions, determined by SmartStation, with angles and distance measured by the integrated total station. This technique takes advantage from several RTK positions that could be available during multiple setups and can eliminate the need of two controls to start the survey. As a result, this technique promises more accuracy and flesibility when working with SmartStation. CUPT was tested on multiple data collected ('5' RTK points and total station observations -forward and backward- for '28' testing points) using a SmartStation and a prism set on a rod. CUPT results shown that, the SmartStation plan coordinates were improved from about '0.05m' RMSE (Root Mean Squared Error) to about '0.014m' RMSE and to better than '0.006m' RMSE for the forward and backward observations respectively. Likewise, the height component was improved from about '0.017m' RMSE to better than '0.005m' RMSE. More improvement was offered by CUPT by combining both of the forward and backward observations, about '0.003m' RMSE in plan coordinates. Other resultsfor various tests are presented in this paper as well.