Industrial Automatic Control Systems and Controllers Annotation << Back Working out of a method of synthesis multiple parametre neuroforecast identifications with use of ensemble of neural networks with indistinct logic for difficult technological installations V.M. BUYANKIN In article the synthesis method identifications with use of neuroforecast 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 нейропрогнозирование transitive characteristics on some steps forward. Keywords: neuropredictive identification synthesis technique, recurrent neuron network ensemble, hybrid neuron network Contacts: E-mail: Viktor-Buyankin@yandex.ru Pp. 31-35.