DETERMINATION OF TRANSMISSION ROUTES IN THE NETWORK OF THE INFORMATION AND TELECOMMUNICATION SYSTEM OF RAILWAY TRANSPORT USING GWO

Abstract

In this work, the study of the possibility of using GWO to determine the routers for transmitting control messages in the network of the information and telecommunication system (ITS) of railway transport is carried out. At the present stage, the OSPF protocol is used in the computer networks of railway transport, when using which a problem appears in real time due to changes in data volumes, and for the solution of which it is advisable to use artificial intelligence methods, which confirms the relevance of the topic. To determine routes in the ITS network of railway transport, it is possible to use both neural networks (Hopfield network, Boltzmann machine, multilayer perceptron, RBF network, neurofuzzy network) and multi-agent methods of intelligent optimization (ant and bee algorithms, the bat algorithm and the grey wolves algorithm). To determine the routes for transmitting control messages in the ITS network of railway transport (at the backbone level), it was created using the Python language and the following libraries: NumPy; NetworkX; Matplotlib; Tkinter software model “Routes_GWO”, which is based on the implementation of Grey Wolf Optimizer with the main parameters: wolf population – 100; the maximum number of iterations is 50. On the created software model “Routes_GWO”, a study of the fitness function by iterations was conducted (from 0 to 50). A series of experiments was organized on the created software model “Routes_GWO”; all the results obtained provide correct solutions result (building a minimal spanning tree), but take much less time than finding a similar solution using a multilayer neural network, which requires determining its optimal parameters (number of hidden neurons, type of neuronal activation function, learning algorithm), creating various samples and organizing the main stages of the neural network: training; testing; validation.