By means of numerical simulation and intelligent algorithms, for thin-walled forgings with vertical and horizontal high ribs of aluminium alloy, taking the ratio of blank height to width, the blank temperature, the forming speed and the friction coefficient in the forming process as the design variables, and taking the filling performance, the organization streamline and the forming uniformity as the optimization objectives, the highly nonlinear relationships between the design variables and the optimization objectives were constructed by the Kriging model. Then, the Kriging model was optimized to obtain Pareto solution set by the multi-objective genetic algorithm (NSGA-II). Furthermore, considering the actual process conditions, a group of relatively balanced optimal solutions were selected for the trial production. The results show that the optimized results significantly improve strain uniformity, filling performance and streamline distribution.