부산대학교 도서관

The alarm system is an overarching component in effectively managing abnormal situations. To enhance its efficacy, we have previously developed an alarm graph modelling (AGM) that overcomes the challenges of traditional methods in alarm workshops. This paper develops an integrated platform that is used for the justification and verification of AGM and other theoretical purposes in the future. Despite the limited availability of process simulations, we successfully updated the Tennessee Eastman Process Simulink model to work seamlessly with the alarm system. The outcomes of the alarm workshop were then implemented, and the alarms were adjusted accordingly. By analyzing predefined scenarios, we were able to evaluate the AGM’s ability to accurately represent abnormal event progression, thereby validating and fine-tuning the alarm system.