Estimation of a plant operator's cognitive state

Recently, plant systems have remarkably improved and their operation has become difficult, whereas there is a shortage of expert operators in Japan. Plant operation supporting systems are required to cope with that situation. Until now such systems did not consider the operator's cognitive state, so that they might bring about human errors. This paper suggests that the plant operation supporting system should estimate the operator's cognitive state, and describes a model to estimate the operator's cognitive states, including an inner translation model and rules to stimulate the model depending on plant process indicators. These models are represented by a fuzzy modeling method. The estimation reasoning is calculated by the Dempster's combination method to integrate matching rules. An Individual model is built experimentally, and usefulness of the presented method is verified.<<ETX>>