Тихонов, В. А.Карташов, В. М.Леонидов, В. И.Тимошенко, Н. В.Рыбников, Н. В.Селезнев, И. С.Олейников, В. Н.2021-06-102021-06-102020Detection-Recognition of Unmanned Aerial Vehicles Using the Composed Auto-Regression Model of Their Acoustic Radiation / В. А. Тихонов, В. М. Карташов, В. Н. Олейников, В. И. Леонидов, Л. П. Тимошенко, И. С. Селезнев, Н. В. Рыбников // Visnyk NTUU KPI Seriia - Radiotekhnika Radioaparatobuduvannia. – 2020. – № 81.– pp. 38-46.doi.org/10.20535/RADAP.2020.81.38-46https://openarchive.nure.ua/handle/document/16409When solving the actual task of detecting by its own acoustic emission (AE) unmanned aerial vehicles (UAVs), making a potential threat to various areas of human activity, it becomes necessary to distinguish its signal from all other acoustic noises. The application of the autoregression model, widely used in practice, is complicated by the need to use significantly high orders of the model, since the distinguishing features of the UAV acoustic signal that differ it from other signals are located in the low-frequency region of the spectrum. The article proposes the use of a composite autoregression model that adequately describes the correlation properties of a signal at significant time intervals and provides an increase in spectral resolution in the low-frequency region. Experimental studies carried out on using the proposed mathematical model show significant differences in the spectral power density (SPD) of UAVs AE from SPD of various sources’ noise, which improve the quality characteristics of the UAV detection-recognition problem. A simplified procedure is proposed for determining frequencies of SPD peaks of a long-term autoregression model without spectrum calculation, which is advisable to use when working in real time.enrecognitionacoustic signalunmanned aerial vehicleautoregressive modelpower spectral densitydetectionArticle