Improving the accuracy of the UAV following along a given trajectory under wind loads by the SPSA method

Abstract One of the main areas of UAV application is the collection of data with their exact binding to the selected coordinate system (for example, aero photography). In this case, it is important to not only get the exact coordinates of data collection, but also to ensure the minimum deviation of the UAV from the path under the conditions of external disturbances (wind loads) acting on it. In the article a procedure for assessing wind speed and direction using the SPSA method is proposed. The results of simulation modeling of the algorithm's operation confirmed during field tests on an ultralight UAV are presented. Key words UAV control system, randomized algorithms, Kalman filter, GNSS, random process prediction methods. Acknowledgements This research is supported by St. Petersburg State University, project no. 73555239.