For the defect problem in deep drawing process of gray hopper, the unidirectional tensile experiment of specimens was carried out based on image method, and the real mechanical property parameters of material were obtained. Then, the forming process was numerically simulated by software Dynaform. Furthermore, taking stamping speed V, blank holding force F and friction factor C as the factors affecting the forming quality of sheet metal, and taking the maximum thinning rate y1 and the maximum thickening rate y2 as the optimization objectives, the response surface method (RSM) experiment was designed by Box-Behnken (BBD), and the quadratic polynomial response objective function between process parameters and y1 and y2 was established to obtain the optimal process parameter combination that affected the forming quality of the gray hopper. Finally, the experimental results show that the scheme can effectively improve the forming quality of the gray hopper, eliminate the forming defects of the gray hopper, and provide an effective reference on the defect optimization for this kind of deep drawing parts which was verified by the numerical simulation and actual production.