Kuzmenko N. Multiparametric data recovering in unmanned aerial system with multi-choice classification of flight situations

The thesis is dedicated to problems of missing UAS flight data recovering bases and multi-choice classification of situations that characterize the deviation of UAV parameters beyond the permitted values according to the degree of danger. The method of multiparameteric recovering of missing UAS data that uses available flight information to fill data gaps and takes into account the measurement errors was developed. Further development of the multi-choice classification of UAV flight situations to analyze unplanned deviations from the set trajectory that takes into account the recovered UAS data by proposed method was done. An improved method for evaluating the probability of an existing flight situation class on the basis of UAV parameters deviations calculation from the planned values and classes metrics was developed. A verification of developed methods by computer simulation and experimental tests was done.