Regularization of the Problem of Monitoring the States of Group Objects of Flight Tests

The article discusses the problem of monitoring the technical condition of prototypes of multi-agent systems. The classical control procedure for multi-agent systems, a new type of group test objects, is analyzed. The fulfillment of the necessary control condition, that is, the observability of the states of the test object for a multi-agent system, is required to ensure a sufficiently high probability of the correct detection of each element of the object. Otherwise, the control result becomes unreliable, because the elements of the vector of measured parameters are mixed. The states of the elements of a group object act as such parameters. To ensure required observability and, consequently, correctness of the control problem, the use of additional highly informative features as regularizers is proposed. The search for these signs is performed in three directions: analysis of the hyperspectral image of elements via the search for unique forms of the spectrum corresponding to the characteristic of a particular element of the material; analysis of the location of these materials and the possibility of various combinations of features of the spectrum shape and the location of materials; and analysis of the infrared portrait of elements in the middle infrared range, in which characteristic bright areas can be distinguished, corresponding to the functional equipment location. The use of these features in terms of processing data from information and measuring systems requires some preparation and, preferably, automation. For automation, it is proposed to use single-pass neural network detectors. The results will be useful in developing a system for information collection and analysis for testing prototypes of multi-agent systems.