AUTOMATION OF THE PROCESS OF MONITORING AND ANALYSIS OF THE EXTERNAL ELECTROMAGNETIC ENVIRONMENT BASED ON THE USE OF A DETECTOR ARTIFICIAL NEURAL NETWORK

Abstract

The article deals with the application of the detector neural network (DNN) for recognition of the parameters of the front of the electromagnetic pulse of a lightning discharge in the automated system for monitoring and analyzing the external electromagnetic environment. DNN allows not only to isolate the pulse front in the structure of the oscillographic image, but also to memorize the measured parameters of the recognized image segment in the concept of the neuron-detector as a result of emulation of an active perceptual act of image perception and emulation of a counter-learning procedure. This makes it possible to reduce measurement errors and increase the speed of signal processing in comparison with the method of interactive analysis of electromagnetic interference characteristics.