Synergistic Neural Network and Velocity Pausing Particle Swarm Optimization for Enhanced Residential Building Energy Efficiency: A Case Study in Kuwait

The global energy demand of buildings is on the rise, driven by factors such as rapid population growth, increasing comfort, technological advances, and ongoing developments in building construction. This escalating energy consumption in buildings is a major contributor to the energy crisis and climate change. Accurate prediction of building energy consumption is essential for gaining insight into energy utilization, reducing waste, and enhancing comfort conditions. This study aimed to introduce a reliable technique for predicting and optimizing the energy consumption of residential buildings, with a focus on a case study in Kuwait. A robust Artificial Neural Network (ANN) was developed, meticulously trained, and rigorously tested to provide accurate energy consumption predictions. Subsequently, an innovative variant of the Velocity Pausing Particle Swarm Optimization (VPPSO) algorithm was employed to identify optimal energy consumption solutions. This novel optimization technique can achieve significant reductions in building energy consumption, with potential savings of up to 43%. Additionally, a sensitivity analysis was performed using the Garson method to assess the impact of input parameters on energy utilization. The results reveal that the insulation and cooling setpoint exert the greatest influence on the objective function, followed by the outdoor airflow. The proposed model, which combines the power of ANN with VPPSO, can be applied to similar buildings, offering precise predictions and optimizing energy consumption. This approach holds promise in addressing the pressing challenges of energy efficiency in building environments.