Optimization of hybrid GA-PSO-based energy management with Six Sigma penalty in buildings

Abstract

This study proposes the optimization of a hybrid genetic algorithm-particle swarm optimization (GA-PSO) building energy management combined with Six Sigma for quality control. The main problems include high energy consumption, large carbon emissions, and performance variability. Six Sigma is applied through control limits (UCL/LCL) and process capability index (Cpk) so that the solution is not only efficient but also stable. Using 30 days of operational data, the model evaluates daily energy consumption (kWh) and carbon emissions, then compares the baseline with pure GA, pure PSO, and GA-PSO+Six Sigma. The results show that GA-PSO reduces average energy consumption by 5.1% compared to GA and 3.5% compared to PSO. When combined with Six Sigma, the savings increased to 7.5% compared to GA and 6.6% compared to PSO, while reducing carbon emissions without compromising operational comfort. These findings present a measurable, sustainable, low-carbon building energy management model that is aligned with the decarbonization framework and ISO 50001 best practices.