LIFT COEFFICIENT IDENTIFICATION OF FLYING VEHICLE DURING THE TAKE-OFF STAGE BY USING THE SPIKING NEURAL NETWORK ACCORDING TO THE NEURAL IZHIKEVICH MODEL AND DEEP LEARNING SPIKEPROP ALGORITHM

Abstract

This article proposes a method to identify the lift coefficient of a flying vehicle during the take-off stage based on the data recorded from the actual flight, using the spiking neural network (SNN) according to the Izhikevich neural model and the spiking error backpropagation algorithm (SpikeProp). The obtained results are compared with the identified results when using the Radial Basic Network (RBN) and the identification by nonlinear regression model (NARX), showing higher accuracy, reliability and number of network training times less. The obtained results serve as a basis for applying SNN models and synthesizing other network training algorithms in identifying aerodynamic coefficients of a flying vehicle in different maneuvering modes.