Browsing by Author "Van der Merwe, V."
Now showing 1 - 1 of 1
Results Per Page
Sort Options
- ItemBenefits of GPS-aided altitude estimation in multirotor drones(Stellenbosch : Stellenbosch University, 2021-03) Van der Merwe, V.; Smit, W. J.; Stellenbosch University. Faculty of Engineering. Dept. of Mechanical and Mechatronic Engineering.ENGLISH ABSTRACT: Ongoing research investigates the use of drones for heliostat calibration,however, is limited by the quality of the collected data and thus the positional accuracy of the drone. Evaluation of various drone models identified the altitude to be the least accurate and in most cases was measured using a single barometer, which is prone to sensor drift and disturbances.To address the issue, the study proposed GPS-aided altitude estimation using a Kalman filter, in addition to sensor fusion based on Covariance Inter-section with the testing done via simulations. In addition to testing a GPS and RTK-GPS, the estimation and correction of the barometer drift were evaluated. Lastly, the relationship between the altitude and drift estimation ac-curacy based on differences in the sensor noise ratios were also investigated.The results indicated minimal benefit and effectiveness in the RTK-GPS and GPS cases, thus exhibiting single sensor dominance. The sensor noise ratios illustrate the benefits of GPS-aided altitude estimation, with the improvement in accuracy being largely dependent on the difference in the ratio. Furthermore, a window for optimal benefit was identified between the ratios of 0.6and 4, with the improvement in accuracy rapidly decreasing as the ratio becomes less than unity. The research provides insight regarding the expected accuracy improvementof the altitude estimate for a wide range of sensor ratios, while also illustrating the accuracy and benefits of drift estimation and correction when using GPS altitude measurements and sensor fusion.