Intelligent Microclimate Control: From Reactive Algorithms to Predictive Models

Abstract

The paper discusses the limitations of reactive algorithms for microclimate control in industrial premises and justifies the need for a transition to predictive models. An analysis of traditional and intelligent approaches is carried out, and the mathematical formulation of the microclimate forecasting problem is presented. The use of the NNARX architecture and its combination with fuzzy logic for the development of a hybrid control system is proposed. The advantages of this approach for improving the stability of technological processes and the energy efficiency of production are demonstrated.