Multi-objective optimization of desiccant wheel via analytical model and genetic algorithm

The dehumidification performance and energy consumption of desiccant wheel are affected by many design parameters and operating variables. However, there are few researches concerning multi-objective optimization. Therefore, an optimal design framework combining analytical model, multi-objective optimization and decision -making is proposed. The model was based on the overall heat and mass balances, so the objective functions of the desiccant wheel were derived, and the constraints for the equilibrium of the adsorption and desorption were obtained. Then, the non-dominated sorting genetic algorithm II (NSGA-II) was employed to calculate the Pareto optimal front of the two-objective optimization, and the results were analyzed with psychrometric chart. This solution set includes not only the optimum one with the design method of psychrometric charts, but also includes other better solutions. From the Pareto solutions, the final optimal solution was obtained with the technique for order preference by similarity to ideal solution (TOPSIS) based on four criteria. With the final optimal parameters applied to existing example, the further improvements on the outlet process air humidity ratio and the dehu-midification coefficient of performance were found.

Automated summary

This research proposes an optimal design framework combining analytical model, multi-objective optimization, and decision-making for optimizing the dehumidification performance and energy consumption of a desiccant wheel. An overall heat and mass balance-based model is used to derive the objective functions and constraints for the desiccant wheel. The non-dominated sorting genetic algorithm II (NSGA-II) is employed to calculate the Pareto optimal front of the two-objective optimization, and the results are analyzed using a psychrometric chart. The final optimal solution is obtained using the technique for order preference by similarity to ideal solution (TOPSIS) based on four criteria, resulting in further improvements on the outlet process air humidity ratio and the dehumidification coefficient of performance when applied to an existing example.