Comparative Analysis of Neural Network Architectures for Intelligent Microclimate Control in Production

Abstract

A comparative analysis of neural network architectures (MLP, RNN, NNARX) for predicting microclimate parameters in industrial cyber-physical systems has been carried out. The advantages of NNARX in reproducing environmental dynamics are demonstrated, and its application for intelligent control is substantiated.