APPLICATION OF NEURONIC NETWORKS FOR CONSTRUCTING ADAPTIVE CONTROL SYSTEMS OF TECHNOLOGICAL PROCESSES

The work covers the neuronic networks (synergetical model, its abilities to structural identification of technological processes). The aim is to study the neuronic networks as a base for the structural identifiers in adaptive control systems. The analysis of exsistence and stability for stationary solutions in the neuronics network has been performed firstrly. The conditions determining the phase space structure have been revealed. The new model of the pattern classification has been proposed, the methods for structural identification of the technological processes on base of the neuronic networks have been proposed. The subsystem for structural identification of the technological processes has been constructed. The facilities permitting to make such subsystems with required properties have been proposed. The software-hardware complex for structural identification of the optic fibre draft has been constructed. The software-hardware complex for identification of the optic fibre draft has been introduced in the Special Design Office of Cable Industry, the software of the individual identification has been introduced in the Internal Affairs Board of c. TolyattiAvailable from VNTIC / VNTIC - Scientific & Technical Information Centre of RussiaSIGLERURussian Federatio