A REDUNDANT GNSS-INS LOW-COST UAV NAVIGATION SOLUTION FOR PROFESSIONAL APPLICATIONS

Abstract. This paper presents the current results for the FP7 GINSEC project. Its goal is to build a pre-commercial prototype of a low-cost, accurate and reliable system for the professional UAV market. Low-cost, in this context, stands for the use of sensors in the most affordable segment of the market, especially MEMS IMUs and GNSS receivers. Reliability applies to the ability of the autopilot to cope with situations where unfavourable GNSS reception conditions or strong electromagnetic fields make the computation of the position and / or attitude of the UAV difficult. Professional and accurate mean that, at least using post-processing techniques as PPP, it will be possible to reach cm-level precisions that open the door to a range of applications demanding high levels of quality in positioning, as precision agriculture or mapping. To achieve such goal, a rigorous sensor error modelling approach, the use of redundant IMUs and a dual-GNSS receiver setup, together with close-coupling techniques and an extended Kalman filter with self-analysis capabilities have been used. Although the project is not yet complete, the results obtained up to now prove the feasibility of the aforementioned goal, especially in those aspects related to position determination. Research work is still undergoing to estimate the heading using a dual-GNNS receiver setup; preliminary results prove the validity of this approach for relatively long baselines, although positive results are expected when these are shorter than 1 m – which is a necessary requisite for small-sized UAVs.