Adaptive Interval Type 2 Neuro Fuzzy Model for Occupant Comfort Prediction and Energy-Efficient Smart Home Control

Abstract

This paper investigates an adaptive type 2 neuro fuzzy control approach for energy efficient smart home IoT systems.  The proposed method integrates a neural network based comfort prediction model, an adaptive type 2 fuzzy(T2F) inference mechanism to generate intelligent control actions for HVAC, lighting, and ventilation systems. Environmental parameters including temperature, humidity, illumination, and occupancy are continuously monitored through IoT sensors and utilized for real time decision making. The adaptive T2F system effectively handles uncertainty and imprecision. Experimental evaluation on smart home environmental data demonstrates improved comfort maintenance, enhanced adaptability, and reduced energy consumption compared with conventional threshold based, classical fuzzy, and neuro fuzzy control approaches. The developed controller can be applied to smart home automation systems because its rule base remains interpretable while adapting to changing environmental conditions, making it applicable to energy management systems.