Sequence Compression and Alignment-Based Process Alarm Prediction

With the increasing complexity ofproduction technologies,alarmmanagement becomes more and more important in industrial process control.The overall safety of the plant relies heavily on the situation-awareresponse time of the staff. This kind of awareness has to be supportedby a state-of-the-art alarm management system, which requires broadand up-to-date process-relevant knowledge. The proposed method providesa solution when such information is not fully available. With theutilization of machine learning algorithms, a real-time event scenarioprediction can be gained by comparing the frequent event patternsextracted from historical event-log data with the actual online datastream. This study discusses an integrated solution, which combinessequence compression and sequence alignment to predict the most probablealarm progression. The effectiveness and limitations of the proposedmethod are tested using the data of an industrial delayed-coker plant.The results confirm that the presented parameter-free method identifiesthe characteristic patterns operational states and theirprogression with high confidence in real time, suggesting it for awider adoption for sequence analysis.