CORMSIS Seminar Event

Abstract: Many engineering systems are designed and deployed to support the accomplishment of some critical missions. For example, the space flight telemetry, tracking, and control (TT&C) system is one of such systems. Mission reliability of a system is the probability the system successfully accomplish a stated mission in a specified environment. To avoid the risk of mission failure, it is necessary to build models and evaluate mission reliability in system engineering.

Researchers have done many works on reliability modeling and evaluation of mission systems, especially on phased mission systems(PMS). Existing methods include combinatorial approach, state space based approach, and simulation approach. However, many challenges still exist in both theoretical research and engineering application.

In this talk, I will introduce our related research efforts on mission reliability modeling and analysis, focused particularly on our two latest works.

For reliability modeling of complex mission systems, an extended object-oriented Petri network (EOOPN) is proposed. Compared with existing Petri net models, EOOPN introduces logic transitions and broadcast places to improve the modularity and re-usability of the model. Logic transitions extended the traditional transitions to describe the logic relations more visually concise and direct. Using broadcast places to share information, EOOPN can overcome the difficulty of constructing a visually complicated network model when the dynamic behaviors of the system's components are highly dependent. The proposed model has been used for reliability modeling of phased-mission systems with common cause failures.

In engineering applications, a mission must be successfully accomplished within a given time interval. In existing literature, it is commonly assumed that for the mission to succeed, the system must remain operational throughout the whole time of the given time interval. However, there are some cases that such assumption is not true. To evaluate mission reliability more accurately for mission systems with time redundancy in mission execution, we identify two cases which have not been studied in existing literature. One is the case that the mission success requires the system keep operational continuously for a minimum length of time during given mission interval, another is the case that the mission success requires the total length of system operational time be greater than a given value during mission interval. By using renewal theory, we derive matrix integral equations for evaluating mission reliability in these two cases. In addition, we give a numerical algorithms for their evaluation. All the results are verified by simulation with numerical examples (this is a joint work with FRSE Prof. Jane Hillston at the University of Edinburgh)