Optimizing multi-chiller energy efficiency using a hybrid genetic algorithm-barnacle mating optimizer (GA-BMO) and dynamic scheduling

Energy consumption in heating, ventilation, and air conditioning (HVAC) systems, particularly chillers, represents a major portion of operating expenditure in large-scale facilities. This study proposes a hybrid optimization-based control strategy that integrates a Genetic Algorithm (GA) with the Barnacle Mating Optimizer (BMO), referred to as GA-BMO, to minimize total chiller power consumption while satisfying cooling demand constraints. The case study is a three-chiller plant inspired by the Malaysian Nuclear Agency configuration with installed capacities of 370 RT, 370 RT, and 340 RT (total 1080 RT). The research was conducted in three stages: algorithm development, comparative performance evaluation, and simulated real-time validation using a MATLAB/Simulink dynamic scheduler integrated with a cloud-synchronized monitoring dashboard. Across six operating scenarios (70%-95% of plant capacity), GA-BMO optimized the partial load ratios (PLRs) and achieved power reductions of 25.37-52.31 kW (3.55%-9.01%) relative to the default baseline operation (static schedule derived from the existing plant settings). The proposed method consistently outperformed standalone GA and BMO as well as benchmark algorithms including Improved Genetic Algorithm (I-GA), Imperialist Competitive Algorithm (ICA), Particle Swarm Optimization (PSO), and Cuckoo Search (CS). Economic analysis using the Non-Domestic Medium Voltage tariff (ToU peak rate) indicated projected annual cost savings of RM 18,587.10-RM 38,361.26 depending on load conditions. Overall, the GA-BMO strategy provides a practical and scalable solution for intelligent chiller scheduling by bridging optimization-based control and real-time-capable monitoring in HVAC energy management.