Temporal reasoning under uncertainty for industrial diagnosis

Industrial applications of artificial intelligence require the ability to manage uncertainty and ream about changes in time. In this paper, we develop an approach for representing temporal systems under uncertainty called state and event time net. The state and event time net can be used to recognize the significance of events and state variables with respect to current plant conditions and predict the hiture propagation of disturbances. In this approach, we allow two kinds of dependencies: causal and temporal. it represents causal dependencies between date variables and temporal dependencies between events. Causal dependencies in the not are represented by probability distributions for the occurrence of events and temporal dependencies are be represented using hizzy linguistic terms. The construction of the net is based on expert knowledge. We illustrate its applicability in do fossil power plant domain. A complementary god of this work is to analyze different research efforts for temporal reasoning under uncertainty using probabilistic and fuzzy logic models.