Fuzzy cognitive map based situation assessment for coastal surveillance

This paper presents a fuzzy cognitive map (FCM) approach to develop an inference engine to perform goal reasoning based on information gathered by a sensor network. Fuzzy nodes are used to represent the goals and sub-goals in the system and conditional causality regulates inference in the network. Statistical estimates obtained from a preceding data fusion operation are converted into fuzzy memberships and then propagated through the causal structure of the FCM. Sensor information based decisions are fused into higher level goals and premises until the system level objective is achieved. The developed situation assessment framework has been evaluated for two search scenario simulations.