Synthesis method neurofforecasting difficult technological installations with use of indistinct intellectual technologies

Abstract

In article the synthesis method neurofforecasting identifications with use of ensemble of hybrid neural networks with indistinct logic is considered at incomplete, inexact static and dynamic characteristics of difficult technological installations. For realisation of this method the block diagramme consisting of ensembles of recurrent neural networks and predicting hybrid neural networks with indistinct logic is offered. The first ensemble of neural networks identifies transitive characteristics of the difficult technological installation, the second ensemble of neural networks spends neurofforecasting transitive characteristics on some steps forward.