The temperature of the cell has a great influence on the performance of the power battery pack, in order to reduce the maximum temperature and improve the energy density of a liquid-cooled power battery pack under high rate discharge condition, the cooling structure parameters of the battery pack module were optimized. Firstly, the electricity-heat model of single cell was established and calibrated, built the battery pack calculation model based on the cell electricity-heat coupling model. Then, the cell spacing and the inlet temperature of coolant were taken as the optimization variables, the maximum temperature and volume of the battery module were taken as the optimization objectives, the maximum temperature difference, the cell spacing and the inlet temperature of coolant were taken as constraints. Used the Latin hypertaxis method to establish a parameterized combination sample for the optimized variables, combined with the Kriging model and multi-objective genetic algorithm for the optimal solution. The optimization results showed that the maximum temperature decreased by 9.7%, the maximum temperature difference decreased by 12.5%, and the cell spacing decreased by 7.1%. Finally, according to the optimization results, the prototype was trial-produced and the bench test was completed, experimental test values and calculated results have a good consistency, so as to verify the reliability of the results and which was conducive to improving the heat dissipation performance and energy density of the battery pack.