A Data-Driven Stochastic Optimization Model for Energy Conservation of Industrial Air Conditioning Systems

Optimization control for energy conservation adapting to environmental dynamics in manufacturing systems is critical. A novel data-driven stochastic optimization control model of an air conditioning system against outside air temperature and internal factory heat dynamics is presented in the article. The proposed model releases the limitation of existing make-up air unit (MAU) control methods present in environmental dynamics. The solenoid valve opening of MAU is optimized to simultaneously minimize energy usage and temperature damping. In addition to taking advantage of heat transfer rules, the proposed model is evolved by a data-driven method. A genetic algorithm coupled with a scenario-based sample average approximation solution procedure is developed to resolve the addressed problem with high complexity. The proposed algorithm demonstrated superior performance compared to the other tested methods. Its effectiveness was particularly significant in scenarios involving large temperature deviations, achieving energy savings ranging from 2.81% to 18.64%. The proposed model and solution method render a promising alternative for contemporary energy conservation paradigms while successfully resolving internal and external temperature uncertainties.