Human Model-Based Dynamic Evaluation for Alarm System in Chemical Plant

In this paper, we propose an operator model to detect and identify causes of failure in an emergency, based on which an alarm system in a process can be evaluated and improved. The operator model is a human processor model that includes a perceptual processor, short-term and long-term memories, a cognitive processor, and a motor processor. Knowledge bases for variable information, failure-symptom relation, and alarm management, as well as an abnormal state supervising procedure, are constructed. Since the knowledge bases are built based on general knowledge of a plant system, the model is easily augmented. Consequently, the operator model substitutes for an operator as a virtual subject to supervise the plant system. The operator model automatically produces a track of fault detection and identification (FDI) in an emergency after a malfunction occurs. By analyzing the FDI tracks for an alarm system, we can evaluate the effectiveness of the alarm system. Case study shows the usefulness of the model-based dynamic evaluation method.