Investigation of Precise Relative Positioning through Varying Equipment Grades

Abstract

In this thesis, three methodologies are investigated in order to provide precise relative positioning knowledge between two dynamic platforms as equipment grade is varied. Two methods are integrated into the real-time kinematic (RTK) algorithm using differential GPS techniques to aid the ambiguity resolution of static and dynamic baselines. Lastly, with the introduction of modern GNSS signals, the benefits of integrating single-frequency (SF) observables from GPS, Galileo (GAL), and the BeiDou (BDS) constellations into a single RTK algorithm is explored. The first method uses an adaptive extended Kalman Filter (EKF) to estimate stochastic properties of single-differenced (SD) GPS combinations. This technique improves the resolution of the carrier-phase ambiguities allowing for precise relative navigation and improved time-to-first fix (TTFF). Secondly, a tightly-coupled RTK algorithm is demonstrated which combines ultra-wideband radio (UWB) observables with SD GPS combinations. This is shown to improve TTFF and increase the robustness of the fixed integer solution. An overview of the estimation techniques is provided, and errors observed in diagnostic assessment tools are explained. To better evaluate the robustness of the presented algorithms, they are applied to experimental data collected with equipment of varying grade. Survey-grade equipment is heavily used in RTK research or in applications with a need for precise relative positioning between a base and rover platform. This equipment can be costly and not applicable to many emerging modular technologies. Low-cost sensor suites have been shown to create noisier observables due to the instabilities of their internal oscillators. In addition, low-cost antennas exhibit irregular gain patterns and poor multi-path suppression which obscure the ambiguity search space leading to longer TTFF and higher chances of incorrectly fixing integers. Thus, it is of interest to evaluate the effects of equipment grade on the ambiguity search space for on-the-fly ambiguity estimation. The investigation of the search space is first assessed using a zero-baseline test. This test provides insight into the observability of the carrier phase ambiguity since no geometric range is embedded in the observables. The study then continues by evaluating the search space during a static baseline test. Measurement innovations are monitored and a unique integer validation scheme is shown to improve the percentage of correct integer fixes for all utilized equipment. Lastly, the RTK algorithm is extended to consider dynamic baselines under the pretense that both the base and rover platforms are mobile. This breaks several assumptions of the nominal RTK algorithm and allows it to be considered a Dynamic-Real Time Kinematic (DRTK) algorithm.