Priority classification of combat unmanned aerial vehicles based on a comprehensive assessment of the main indicators

In this work, we solved the problem of obtaining estimates of the prospects combat of unmanned aerial vehicles (UAVs) according to the available estimates of a certain part of UAVs and their further clustering by the value of the complex priority indicator. A new estimation scheme is proposed that uses five known particular indicators of each assessed UAV, which, after normalizing them and determining the significance coefficients of particular indicators, were combined into a complex UAV indicator. Based on the obtained estimates of the complex indicator of each of the considered UAVs, the clustering of the set of the considered UAVs was carried out according to the belonging of the complex indicator to the intervals of its possible values from 0 to 100. One hundred or zero points can be obtained by that UAV, in which all particular indicators have respectively the best or worst values among all compared UAVs. Clustering made it possible to divide the UAVs into groups according to the principle of similarity of the tasks performed. In the Python programming language, a program was developed that implements the proposed clustering scheme, and allows to obtain estimates of objects and classify these objects according to the principle of proximity of complex indicators. This scheme allows to identify the best (and worst) UAVs in each of the UAV classes, distinguished by the commonly used clustering indicator, takeoff weight. About 14% of UAVs considered in this work were included in the best cluster in terms of complex indicator with an indicator value of at least 59 points with a maximum score of 100 points. The obtained results can be used in the development of requirements for the equipment and operation of UAVs.