Ambiguity resolution with double troposphere parameter restriction for long range reference stations in NRTK System

Abstract

The correct ambiguity resolution between reference stations is the key to calculate the high precision Network Real-Time Kinematic (NRTK) differential information. For long range reference stations (≧50 km), the double difference troposphere model residuals should be considered as the parameters being solved, but this will aggravate the ill conditioning of ambiguity resolution (AR) model between reference stations; as a result, the ambiguity fixing becomes more difficult for the case of long range station ambiguity resolution. In the paper, a new method with double troposphere parameters restriction is put forward for ambiguity resolution of long range reference stations. The proposed method applies GPT2 model, which is called the state of the art empirical troposphere model, to form a high precision troposphere a priori estimation to provide high accuracy double difference troposphere delay estimation. Based on the principles of TIKHONOV regularisation, a regularisation criterion for the double difference restriction model is then built. The difference between the troposphere estimation and the truth value is used as a restriction parameter to improve the estimation of the unknown parameters and optimisation of the ambiguity search range. Trials verify the significant reduction in the ill conditioning of the parametric resolution functions when the double troposphere restriction model is applied. The success rate of ambiguity resolution within 60 s is above 98% for baselines over 80 km, which is an immense improvement from conventional methods.

Keywords:

• Double troposphere restriction
• Network RTK
• Select Power Fitting
• Ill conditioning

Acknowledgements

The authors thank the anonymous reviewers and the editor for their valuable comments on the manuscript. This work is supported by the Key Projects in the National Science & Technology Pillar Program during the Twelfth Five-Year Plan Period (No. 2012BAJ23B01).