To solve the problem of uncertain system contribution under different scenarios for the problem of portfolio optimization, an optimization and decision-making method based on robust capabilities is proposed. First, the quantitative description of the robust capabilities and fuzzy capabilities risk is given. Then, a multi-objective portfolio optimization model is established, balancing maximum capabilities, minimum construction risk and cost, within the constraints of military capability requirements. The improved non-dominated sorting genetic algorithm 3 (NSGA3) is adopted to solve the model by generating a set of non-inferior solutions that meet the decision-making requirements. Finally, simulation experiments verify the effectiveness of the model and algorithm, and can provide support for the overall construction plan of the system.