A rule-based approach for planning counter-actions with multilevel flow modeling

Response planning is one of the most important cognitive tasks when operators deal with anomalies in complex process systems such as nuclear power plants. Operators can face great challenges when no existing procedure can address the current plant situations, as referred to the knowledge-based decision making. It is crucial to develop a support system to aid operators in the decision making of planning. The support system is expected to be capable to identify not only essential resources but also executable solutions. Multilevel flow modeling (MFM) is a functional modeling methodology that can describe planning resources in multiple abstraction levels of systems' intentional knowledge in perspective of functions and goals. In this paper, a rule-based approach for planning counter-actions particularly operating procedures with MFM is proposed. In order to implement the model-based planning procedure, new rules for identifying control actions and their conditions are defined in addition to the existing causal reasoning rules for MFM. Action solutions, i.e. operating procedures to counter the abnormal situation can be identified from a tree structure including all desired states and required actions that are inferred from a prescribed goal by using different categories of reasoning rules.