The invention discloses an NSGAII-based multi-objective optimization decision-making method for a stratospheric aerostat. The method comprises the following steps of (1) building a stratospheric aerostat model, wherein the stratospheric aerostat model is a bielliptical geometric model composed of two semiellipsoids; (2) selecting skin volume, a resistance coefficient and a maximum loop stress of the aerostat as optimization objectives, and selecting skin thickness and a geometric slenderness ratio of the aerostat as optimization variables according to the built aerostat model; (3) performing optimization calculation on the aerostat model by using a multi-objective evolution algorithm, and performing continuous iteration to obtain a non-inferior solution of Pareto front; (4) performing decision-making analysis on an optimization result by adopting a fuzzy entropy weight-based TOPSIS (Technique for Order Preference by Similarity to Ideal Solution) method; and (5) determining an optimal scheme by adopting the TOPSIS method. According to the method, a relatively optimal scheme closest to a positive ideal solution and far away from a negative ideal solution is obtained through the fuzzy entropy weight-based TOPSIS multi-attribute decision-making method.