Abstract: Abstract: A hybrid refrigeration system of refrigerator car was designed in this paper. Photovoltaic energy storage system was the auxiliary power source of the refrigeration system and it drove the refrigerator during refrigerator car parking for a long time midway or at night. Engine drove refrigerator when refrigerator car was driving. The experimental analysis of hybrid refrigeration system of refrigerator car found that power generation of solar panels could only keep electric compressor working for 4.13 hours in one day. The power generation was not enough to support refrigeration system working for all night and this would affect the quality of cargo in the refrigerated carriage. In order to solve this problem, on the one hand, the energy consumption of refrigeration system needed to be reduced. On the other hand, the power generation capacity should be increased. Firstly, mathematical model of the hybrid refrigeration system of refrigerator car was established. Then, experiments about photovoltaic energy storage system and refrigeration system have been down. Correctness of the model was verified by comparing experimental data and simulation data. On this basis, the refrigeration efficiency and economic coefficient of photovoltaic power generation were put forward as the system evaluation indexes in this paper. The refrigeration efficiency was the ratio of energy consumption to cooling capacity and represented the energy consumption of the refrigeration system. If the refrigeration efficiency is lower, the energy saving of the refrigeration system is more. The economic coefficient of photovoltaic power generation is the ratio of power generation of solar panels to the power capacity of batteries and it represents the power generation capacity of per-unit cost of photovoltaic energy storage system. If the economic coefficient of photovoltaic power generation is greater, the power generation capacity per unit cost of photovoltaic power generation system is more. The two evaluation indexes were unified as the objective function. Nine constraints were enumerated. Based on chaotic ant colony algorithm, the hybrid refrigeration system of refrigerator car was optimized. The results after optimization showed that the energy consumption of the refrigeration system was reduced by 238.63 W, the cooling capacity was increased by 103.11 W and the refrigeration efficiency was reduced from 0.542 to 0.402. The optimized refrigeration system was more energy-saving. The volume of optimized refrigeration system was smaller, which made refrigeration system easier to install on the refrigerator car. Also, the power capacity of batteries was decreased by 2.67 kW·h, the whole day power generation capacity of solar panels was increased from 4.45 to 5.82 kW·h and the economic coefficient of photovoltaic power generation was increased from 0.464 to 0.840. In summary, the optimized hybrid refrigeration system of refrigerator car was more energy-efficient and economical. The optimization method and encouraging results in this paper will provide a fundamental reference for optimization design of the research on solar refrigeration system.