THE ALGORITHM FOR DECISSION OF CONTROL SYSTEM SYNTHESIS PROBLEM BY THE ARTIFICIAL NEURAL NETWORKS’ METHOD

Abstract

In the presented article analytical methods of the solution of a task synthesis in a control system are stated. The author considers a neural network of Kokhonen which belongs to the class of networks of direct distribution. It has the only hidden layer of neurons presented in the form of a one-dimensional or two-dimensional lattice. By the author it is presented that the mechanism of cooperation realizes the principle according to which the most excited neuron strengthens (through synoptic weight) not only itself, but also spatially neurons, close to it.Article purpose - analytical methods of the solution of a problem of synthesis in a control system.

Keywords

• синтез управления
• искусственная нейронная сеть
• вариационный генетический алгоритм