Metric Based Design of Integrated Engineering Plants for Robust Performance During Hostile Disruptions

In military applications, it is important for a platform (warship, aircraft, etc.) or an installation (airbase, etc.) to maintain war fighting ability after being damaged. In particular, if the unit requires electric power, cooling, or other resources to perform its mission, then these resources must be available following a weapon detonation event. The integrated engineering plant is responsible for providing these services to the mission critical loads in a unit. This work begins with the description of a notional integrated engineering plant for an all-electric warship. This plant includes ac electric power networks, a zonal dc electric power distribution network, a seawater cooling network, and freshwater cooling loops. A layered modeling approach is used to capture the complex dynamic interdependence of these tightly coupled systems. With this approach, it is possible to examine the propagation of faults and cascading failures throughout the dynamically interdependent systems. Novel continuity of service metrics for integrated engineering plants are set forth. These metrics provide a means of predicting the average and worst case level of service the plant can provide, as well as the worst case scenario. This provides a method of making meaningful comparisons between different designs. The computation and meaning of the proposed metrics are explored using the notional plant. To understand how architectural decisions impact continuity of service early in the design process, it is necessary to solve certain robust design problems. Many robust design problems can be described using minimax optimization problems. A new method of solving minimax optimization problems is proposed. The performance of this algorithm is shown to compare favorably with the existing methods on test problems. Finally, the proposed minimax optimization algorithm is applied to the design of the notional integrated engineering plant. Architecture metrics which allow quantitative comparisons of entire classes of designs are set forth. The spatial layout of the plant is optimized to improve the worst case performance of the plant. Additionally, the valve settings of the seawater cooling system are optimized to achieve optimal worst case performance. It is further demonstrated that the design of the spatial layout and the valve settings can be performed simultaneously. Future research in the area is proposed.