Network Real-Time Kinematic GPS (NRTK) enables almost instantaneous positioning at centimetre level accuracy. A service offering such high precision positioning has many applications, including precision farming, building and property set outs and engineering stake-outs. The number and diversity of these applications is growing as acceptance of the technology increases. The economic benefits of these applications are great. However, the use of NRTK is currently limited to a range of some tens of kilometres from reference stations. The main reason for this limited range is the difficulty in estimating the ionospheric and tropospheric biases at the rover position. The estimation of biases at the rover end, otherwise known as interpolation, can be performed via a variety of techniques. One of the existing interpolation methods is Least Squares Collocation (LSC). LSC requires a functional and a stochastic model to implement the interpolation. This research develops algorithms that enable superior functional and stochastic models to be generated in real-time for tropospheric bias interpolation, with a view to future ionospheric bias interpolation improvement. The improved algorithms result in superior estimates of tropospheric biases at the rover end as well as an accurate quality measure of these biases. The outcomes of the improved ability to interpolate tropospheric biases will be that the distance from the CORS sites that NRTK can be performed will be greater, the accuracy and reliability of NRTK positions will improve and that rover end ambiguity resolution for NRTK will be quicker and more reliable.
展开▼
