An Adaptive Optimization Approach Based on the Human factors and Its Application to Process Alarm Thresholds

Currently, the majority of studies on process alarm thresholds optimizations are less concerned with human factors. In response to this limitation, a novel adaptive optimization approach to process alarm thresholds concerning human factors is explicitly introduced in this paper. Apriori algorithms are used to extract association rules between process alarms and manipulated variables. By means of fuzzy reasoning strategies, a human factor index in terms of operators’ perception and treatment capabilities is created before the nuisance alarm rate and alarm treatment rate are calculated, respectively. Subsequently, an thresholds optimization objective function with respect to the weighted false alarm rate (FAR) and missed alarm rate (MAR) is established, in which, the FAR and MAR are estimated with kernel density methods while weights are calculated based on the human factor index. Case studies show that the proposed method can adapt to operators’ human factors, facilitate the process alarm treatment and improve the process safety.