Alarm management via temporal pattern learning

Industrial plant safety involves integrated management of all the factors that may cause accidents. Process alarm management can be formulated as a pattern recognition problem in which temporal patterns are used to characterize different typical situations, particularly at startup and shutdown stages. In this paper we propose a new approach of alarm management based on a diagnosis process. Assuming the alarms and the actions of the standard operating procedure as discrete events, the diagnosis step relies on situation recognition to provide the operators with relevant information on the failures inducing the alarm flows. The situation recognition is based on chronicle recognition where we propose to use the hybrid causal model of the system and simulations to generate the representative event sequences from which the chronicles are learned using the Heuristic Chronicle Discovery Algorithm Modified (HCDAM). An extension of this algorithm is presented in this article where the expertise knowledge is included as temporal restrictions which are a new input to HCDAM. An illustrative example in the field of petrochemical plants is presented.