The improvement in integer ambiguity resolution with INS aiding for kinematic precise point positioning

摘要

Despite the benefits of integer ambiguity resolution (IAR) in precise point positioning (PPP), observation outages and harsh signal environments still impact float ambiguity estimation in kinematic surveying, consequently resulting in ambiguity-fixed failure. The inertial navigation system (INS) is an autonomous and spontaneous positioning one, which could provide continuous and superior positioning accuracy over short time. Thus, the INS attains more accurate position than code solution. Moreover, the tight integration of INS and PPP is capable of continuous operation where there are less than four satellites available. These advantages can improve float ambiguity estimation and assist in re-initializing the interrupted ambiguity and PPP solution. Based on the good quality of float ambiguity, the ambiguity dilution precision (ADOP) and the size of integer ambiguity search space are reduced, and then, the IAR-PPP is improved. In this work, the INS aiding effect on IAR-PPP was revealed by the sufficient theoretical analysis and performance assessment. A ring laser gyroscope-based navigation-grade IMU and a fiber optic gyroscope-based tactical-grade IMU were utilized to conduct experiments in an open-sky environment and urban area. The assessment adopted the following aspects of ADOP, bootstrapping success rate, time to fix and position errors. It is found that IAR-PPP with INS aiding achieves an enhanced performance during GPS outage when INS could deliver a superior accurate position. For the navigation- and tactical-grade IMU, the INS-aided ambiguity re-fixing performance can be classified as three levels: significant improvement for the outage duration less than 10s, moderate improvement for the outage duration from 10 to 60s and a little or zero improvement for the outage duration longer than 60s. From the viewpoint of the INS-predicted position domain, an accuracy better than 0.1m and 1.0m is required for the significant and moderate improvement, while one can only achieve a little or zero improvement if the position error is larger than 1.0m. Besides, we also performed the INS-aided IAR-PPP in real urban environment. For the urban environments, the span of clean data is often shorter than 30min due to intermittent signal interruptions; thus, ambiguity re-fixing for PPP always fails. INS-aided information could bridge the data gaps and achieve fast ambiguity re-fixing. In summary, INS aiding information is capable of improving IAR-PPP performance significantly over a short GPS outage.