A novel multi-objective global optimization method combining refrigerator dynamic model and Genetic algorithm NSGA-II is proposed for increasing overall performance of household refrigerator. A novel refrigerator with heat-storage condensers and an ordinary refrigerator with conventional hot-wall condensers are optimized by the method for multi-objectives of minimizing total cost and energy consumption per 24 h. After the optimization, overall performances of the novel refrigerator and ordinary refrigerator are both increased. Optimized curves on multi-objectives of the novel refrigerator and ordinary refrigerator are obtained and compared with each other. In the optimized curve, for one value of energy consumption per 24 h, a corresponding optimized total cost is acquired. Optimization result of the novel refrigerator is better than that of the ordinary refrigerator. Under the condition of same total cost, energy saving of optimized novel refrigerator compared with optimized ordinary refrigerator is from 20% to 26%. Under the condition of same energy consumption per 24 h, cost saving of optimized novel refrigerator compared with optimized ordinary refrigerator is from $1.8 to $3.4. Crown Copyright (C) 2014 Published by Elsevier Ltd. All rights reserved.