For an automobile reinforcement part, the corresponding data between hot stamping process and forming index were obtained by finite element numerical simulation method. In addition, the highly nonlinear mapping relationships between design variables and design indicators were constructed by the Kriging model, and the global optimal parameters were obtained by using genetic algorithm to optimize the Kriging model. The optimization strategy and optimization results show that the numerical simulation results are in good agreement with the experimental results, and the finite element analysis method can successfully capture the deformation mode and characteristics of the hot stamping process for plate. Therefore, the combination of Latin hypercube sampling and Kriging model accurately predicts the forming quality of products and is applied to the design of hot stamping temperature and blank holder force. Furthermore, the combination of genetic algorithm and Kriging model improves the forming quality of an automobile reinforcement part significantly and eliminates the crack defects of sheet metal in the hot stamping process before optimization. Thus, the feasibility of optimal parameters is verified by experiments, and the optimization strategy has good applicability for the hot stamping process optimization of ultra-high strength steel plate.