Intelligent Lighting Control: A Neuro-Fuzzy Approach for Optimal Energy Efficiency in Smart Homes

This paper presents an Adaptive Neuro-Fuzzy Inference System (ANFIS) specifically designed for an in-telligent control system for domestic lighting in a smart home environment. The aim is to create an efficient model that integrates both fuzzy logic and a neural network to improve the adaptability of the control system. To obtain this type of control, two key inputs are used: the indoor lighting level and the presence of a person. These factors are essential to determine the behaviour of the lighting system in different situations. The ANFIS model is trained using two different learning methods: backpropagation and a hybrid technique. To improve accuracy, eight different membership functions (MFs) are used in three sets of epochs: 20, 50, and 100. The training results give us the best choice of ANFIS model parameters that guarantee the desired performance and minimize errors. The results of the study show that the application of the present approach to lighting systems demonstrates the effectiveness of autonomous control.