Artificial Neural Network Based Energy Optimization For Smart Building Energy Management Systems In Commercial Building

Abstract

As the population grows and human needs increase, so does the number of buildings, leading to higher energy demand and challenges for sustainable use. Therefore, it is essential to develop strategies to reduce energy consumption. Previous studies indicate that poor control over energy usage in commercial buildings has resulted in significant energy wastage. To address this issue, energy management systems have been developed in recent years, playing a critical role in optimizing energy use in various building types. The PID controller, known for its simplicity and ease of use, has become the standard controller in many energy management systems. However, its performance limitations in complex tasks and the overshoot problem in system output reduce overall energy efficiency. This paper proposes an Artificial Neural Network (ANN) Controller to replace the PID Controller for managing and monitoring HVAC and lighting systems, which are the two largest energy consumers in commercial buildings. The controllers were designed and implemented using MATLAB Simulink, where their performance was evaluated under different operating conditions, including step changes and ramp behaviours. The ANN Controller was compared to the PID Controller to evaluate its effectiveness in energy management. In the HVAC system, it eliminates overshoot at 15°C, whereas the PID Controller shows 41.00%. For the lighting system, it achieves zero overshoot at 250 Lux, while the PID Controller reaches 26.67%. These findings confirm its effectiveness in enhancing energy efficiency in commercial buildings by minimizing energy fluctuations and improving system stability, making it a reliable solution for optimizing energy use.