Abstract:

In view of the difficulties in dynamic reconfiguration and resilience evaluation faced by modern combat system-of-systems in a highly adversarial environment, and the deficiencies of existing studies in depicting high-order interaction relationships and cluster evolution mechanisms, this paper proposed a resilience modeling method for combat system-of-systems integrating hypernetwork and game theory, aiming to analyze the resilience mechanism of the system-of-systems in all dimensions from micro, mesoscopic, to macro levels. At the micro level, a high-order motif structure was introduced to represent the complex interaction modes among combat units, which overcame the information loss of traditional binary relationships in depicting multi-element synergy. At the mesoscopic level, a network evolutionary game model based on cumulative payoffs was constructed, and the combat clusters were guided to evolve from individual rationality to cluster optimality through an imitation mechanism to enhance the cooperative reorganization capability. At the macro level, the evolution law of the giant component of the hypernetwork was derived based on percolation theory, and a quantitative evaluation indicator for system-of-systems resilience considering time factors was proposed. The simulation results indicate that compared with the traditional network model, the proposed high-order interaction model exhibits stronger robustness under both random attacks and deliberate attacks; the designed evolutionary game strategy enables the system-of-systems to maintain a component size of over 50% in multiple rounds of confrontation, which is significantly superior to the random recovery strategy. The research results can provide theoretical support for the resilience design and optimization of kill webs of intelligent combat system-of-systems in the future.

Key words: combat system-of-system, system-of-system design, resilience evaluation, hypernetwork, network evolutionary game