Variable Selection for Fault Detection Based on Causal Discovery Methods: Analysis of an Actual Industrial Case

Processes

FabioC. Diehl¹  PedroH. Thompson¹  ThiagoK. Anzai¹  Nayher Clavijo²  Afrânio Melo²  RafaelM. Soares²  MaurícioM. Câmara²  Tiago Lemos²  JoséCarlos Pinto²  LuizFelipe de O. Campos² 

[1] Centro de Pesquisas Leopoldo Americo Miguez de Mello—CENPES, Petrobras—Petróleo Brasileiro SA, Rio de Janeiro CEP 21941-915, RJ, Brazil;Programa de Engenharia Química/COPPE, Universidade Federal do Rio de Janeiro, Cidade Universitária, CP 68502, Rio de Janeiro CEP 21941-972 , RJ, Brazil;

关键词: fault detection and diagnosis;    variable selection;    feature selection;    causality;    conditional mutual information;    real oil and gas process facility;   

DOI  :  10.3390/pr9030544

来源: DOAJ

【 摘 要 】

Variable selection constitutes an essential step to reduce dimensionality and improve performance of fault detection and diagnosis in large scale industrial processes. For this reason, in this paper, variable selection approaches based on causality are proposed and compared, in terms of model adjustment of available data and fault detection performance, with several other filter-based, wrapper-based, and embedded-based variable selection methods. These approaches are applied in a simulated benchmark case and an actual oil and gas industrial case considering four different learning models. The experimental results show that obtained models presented better performance during the fault detection stage when variable selection procedures based on causality were used for purpose of model building.

【 授权许可】

Unknown