Multi-objective optimization of a free-piston Vuilleumier heat pump using a genetic algorithm

摘要

A free-piston Vuilleumier heat pump (FVHP) has unique advantages for residential/commercial heating and cooling applications. This work focuses on optimizing the adjustable machine parameters of a FVHP to maximize its thermal performance for both cold climates, in which the FVHP primarily operates in heating mode, and warm climates, where cooling mode operation dominates. Nine machine parameters are optimized using a thermodynamic model combined with a non-dominated Sorting Genetic Algorithm. The typical optimization time on a desktop PC is several hours. The machine performance is evaluated using a thermal model for a FVHP for the most favorable designs, with the optimized solutions found to provide a 20-30% improvement in output, COP and parasitic energy use compared to the parameters from a current prototype that serves as the baseline. The approach developed should be applicable to most FVHP designs and related devices, such as free-piston Stirling engines.