Clusterização e classificação de alarmes industriais utilizando word embeddings

Abstract

The analysis of industrial alarm content is crucial for detecting and preventing failures in operational processes. Alarms act as an alert system, signaling the operations team about abnormal conditions and potential failures in real-time. However, the excessive generation of records by these systems can make it difficult to identify and respond effectively to critical situations. Therefore, it is essential to develop efficient alarm management strategies to prioritize and group them intelligently. Furthermore, by conducting a thorough analysis of industrial alarm data, it is possible to gain a deeper understanding of operational conditions, recognize recurring patterns, identify trends, and take proactive measures to prevent failures. In this study, data from events and alarms were collected from the SCADA system of a thermoelectric plant located in the state of Paraíba. Natural language processing (NLP) techniques were used for the preprocessing of alarm texts, allowing for the generalization of information by excluding equipment identifiers and low-relevance semantic words. The BERT language model was used for the numerical representation of the texts, and clustering and classification techniques were applied for the efficient grouping of records. This approach not only improves alarm management but also contributes to a safer and more efficient operational environment, which is essential for the industry’s sustainability and productivity.