System-of-Systems Design Optimization Based on Coevolutionary Game

The design optimization of the system of systems (SoS) is of great significance but extremely difficult. A collaborative optimization method based on coevolutionary game is proposed to solve this problem. More specifically, it first provides a standardized design decision model for a large number of heterogeneous systems within an SoS in order to reduce the organizational complexity of the design process significantly, and then uses different game models to describe the collaborative interactions between heterogeneous systems and effectively deal with various coupling constraints and objective conflicts in the evolutionary design process of the SoS. Then, the multi-attribute utility theory, experimental design method, response surface modeling technology and differential evolution algorithm are combined to find a satisfactory global optimal solution with the least information exchange and iterative process. Finally, the design of the over-the-horizon amphibious operation equipment SoS is used as an illustrative example to demonstrate the effectiveness of the proposed method.