Assessment Model and Optimization Strategies of Resilience for Equipment Systems-of-Systems

The equipment system-of-systems (ESoS) plays a pivotal role in military operations, however, the continuous development and expansion of military confrontations can introduce increased complexity, uncertainty, and vulnerability to the ESoS, particularly when confronted with disruptive events. Although resilience is a pivotal aspect in comprehending and managing the damage and recovery of the ESoS, the current research on the ESoS resilience model, which should encompass comprehensive and effective assessment and optimization, still lacks sufficiency. Hence, this paper proposes a resilience assessment model and explores the optimization of recovery strategies based on this model. Specifically, we use a heterogeneous network model to characterize the ESoS by incorporating diverse types of functional entities and information. Next, developing concepts for operation plans and considering uncertain information, the ESoS combat network's performance metric was formulated for ESoS resilience analysis by using rough set theory. It also develops a resilience assessment model based on network topology and network performance. Then, research on the recovery of the ESoS resilience process, we present an optimization strategy for resilience based on reinforcement learning and six other recovery strategies from different dimensions to determine the recovery sequence and maximum resilience. Finally, the effectiveness of the proposed model and the strategies is demonstrated through resilience analysis of an ESoS combat network model under various conditions, which is built according to the assumed combat scenarios. The results provide valuable insights for designing and building a more robust and resilient ESoS.