The ability to autonomously mitigate damage impact to shipboard systems is a key component to survivability, cost reduction and reduced manning in future platforms. In the context of distributed shipboard systems, the process of automated damage impact reduction entails locating and bypassing damaged sections of distributed transport lines. Approaches have been developed to autonomously locate damaged sections in distributed systems such as cooling water and low pressure air systems. The algorithms employed in these systems open and close valves with the aim identifying and isolating damaged transport sections. The valve switching strategies range from regimented switching sequences to totally random switching sequences. The algorithm parameters have been empirically determined based on the performance observed in experimental prototypes. As part of an attempt to globally characterize these parameters and the impact of valve dwell time, a probabilistic model is considered. The approach entails a simplifying one dimensional projection of a distributed hyper dimensional probabilistic density model. The projection characterizes switching and dwell times in terms of probabilities which relate to the convergence rate.
展开▼
