摘要：As systems continue to grow in scale and complexity and have to operate safely in challenging disruptive environments, system safety and resilience has become a critical requirement. This recognition has drawn attention to the concept of resilience, which has different definitions and several different interpretations that tend to be domain specific. For example, resilience in health care clinics means something quite different than resilience in self-driving cars, or energy grids. This paper reviews the different characterizations of resilience and assesses their value proposition in realizing engineered resilient systems. This paper emphasizes the importance of systems modeling in engineering resilient systems and presents an overarching methodology that employs different modeling approaches for operational tasks as a function of problem context. This paper specifically focuses on systems modeling in partially observable and potentially hostile environments. It discusses the need for system model verification, which is key to safety, and system flexibility and adaptability, which are key to resilience. It introduces a formal, probabilistic modeling construct called the “resilience contract.” This construct employs a state-based representation that formalizes the concept of resilience while enabling system model verification and affording requisite flexibility for adaptation and learning. The key findings of our research are that different system modeling approaches and algorithms are needed based on mission tasks and operational context; adaptive capacity and continual adaptability are the two promising characterizations of resilience that can be cost-effectively realized in real-world systems; and the resilience contract construct is an effective means for probabilistic verification of system model correctness while affording flexibility needed for adaptation and learning. Collectively, these findings contribute to the body of knowledge in both model-based systems engineering (MBSE) and engineered resilient systems.
