UDC 621.313.2:681

DEVELOPMENT OF MATHEMATICAL MODEL OBJECT IDENTIFICATION

Veresova Olga Vladimirovna¹, Volkov Vladimir Nikolaevich¹
¹Severstal-Promcervice

Abstract
In the framework of the plan is to develop a universal model identification. The identifier is used by a neural network. For training the network uses the back-propagation algorithm errors. As a dynamic object on which the investigation of the possibility of identification, selected aperiodic link of the second order. The dynamic link moved a random signal with a normal distribution and filmed the response object.

Article reference:
Development of mathematical model object identification // Modern technics and technologies. 2014. № 12 [Electronic journal]. URL: https://technology.snauka.ru/en/2014/12/5204