NEURO-FUZZY MODELING OF A DECISION SUPPORT SYSTEM BY A GROUP OF UNMANNED AERIAL VEHICLES UNDER THE INFLUENCE OF THE EXTERNAL ENVIRONMENT

Abstract

The work discusses the relevance and concept of building an intelligent decision support system based on the use of elements of the theory of fuzzy sets together with the technology of neural networks. The existing requirements for the intellectualization of the control system of a heterogeneous group of unmanned aerial vehicles force developers to switch to the use of artificial intelligence technologies that allow for the training of complex energy-consuming algorithms and ensure their self-organized functioning when solving tasks for their intended purpose. Due to the fact that the process of functioning of a heterogeneous group of unmanned aerial vehicles is described by complex models that take into account the heterogeneity of the group, versatility and multiscenarity when performing tasks for their intended purpose, there is a need to model these processes on a computer. The purpose of the simulation is to obtain a digital version of the double of a real process or object. The probabilistic nature of the description of environmental influences encourages the construction of a model that takes into account all possible behavioral strategies as much as possible and the development of various counteraction scenarios. The most suitable technology capable of analyzing all strategies of the external environment and forming the resulting vector of all situations arising in the process of performing tasks by a group of UAVS for their intended purpose is the well-proven technology of neural networks. However, just getting one solution at the output of the system is not enough, so they resort to modeling situations based on the rule base. Currently, it is easy to form such a base and build a fuzzy model of actions based on it, when the outcomes of alternatives are known inaccurately and the probabilities of their occurrence are estimated using the membership function. Decision-making is based on the method of an ideal point, and the range of acceptable solutions is constructed using a genetic algorithm. The introduction of neural network technology will maximize the viability of the functioning of a group of unmanned aerial vehicles when controlling it both with the help of an operator and offline. The article considers the model of functioning of a group of unmanned aerial vehicles in the conditions of destructive environmental influences.