AHP-RNN Based Hybrid Model with Embedded Software Solution for Smart Climate Control in IoT Automation

The integration of automation through the inter-connectivity and controllability offered by the Internet of Things (IoT) has garnered significant attention in recent years, yet existing solutions often lack the ability to effectively integrate human judgment and predictive analytics. This paper presents a hybrid approach utilizing the Analytical Hierarchy Process (AHP) and Recurrent Neural Network (RNN) to address this gap and achieve smart climate control automation across diverse environments. Our key contributions include the development of an AHP-RNN model that leverages expert judgment to derive decision criteria weights and uses historical data to train an RNN for predictive control. Implementation involves the deployment of temperature, humidity, smoke, and occupancy sensors, with data collected and processed through an online database connected to a common IoT hub. The proposed model demonstrates significant improvements in user satisfaction, energy efficiency, and adaptability compared to existing climate control solutions. Experimental results show that our hybrid approach outperforms state-of-the-art models by achieving higher accuracy in predictive control and better alignment with user preferences.