This paper presents a model based decentralized optimization method for vapor compression refriger- ation cycle (VCC). The overall system optimization problem is formulated and separated into minimizing the energy consumption of three interactive individual subsystems subject to the constraints of hybrid model, mechanical limitations, component interactions, environment conditions and cooling load
demands. Decentralized optimization method from game theory is modified and applied to VCC opti-
mization to obtain the Perato optimal solution under different working conditions. Simulation and
experiment results comparing with traditional oneoff control and genetic algorithm are provided to
show the satisfactory prediction accuracy and practical energy saving effect of the proposed method. For
the working hours, its computation time is steeply reduced to 1% of global optimization algorithm with consuming only 1.05% more energy consumption.