In the aerodynamic drag reduction optimization of automobile tail structure，there is a phenomenon between geometrical characteristic parameters，which makes optimization become blind and complex. In order to ascertain the interaction law of the key geometrical parameters，this paper carried the CFD simulation research of the three main characteristic parameters on the back of the Ahmed model on the basis of the HD-2 wind tunnel experiment. In order to overcome the blindness of optimization，the integrated optimization platform was applied to optimize the tail characteristic parameters. The results show that the angle of the back inclination has the greatest contribution to drag reduction，the two sides rounded corners on the back has the smaller contribution，the contribution of the top corner of the back is the smallest. The influence of the change of three variables on the aerodynamic resistance is non-monotonic. When the values of the angle of the back inclination，the top corner of the back and the two sides rounded corners on the back are 13°，283 mm and 58 mm respectively，the drag reduction rate reaches 11.76%. This paper provides a reference for the study of the optimization of drag reduction in specific models.