Abstract

COLL, Amanda Martí et al. Methodology proposal for fault diagnosis based on decision trees and fuzzy logic. EAC [online]. 2022, vol.43, n.2, pp. 1-16.  Epub Oct 17, 2022. ISSN 1815-5928.

The objective of this research is to present a methodology proposal for the diagnosis of failures in the industry based on decision trees and fuzzy logic. Decision trees have among their advantages simplicity and ease of interpretation. While fuzzy logic provides capabilities for managing the uncertainty of the knowledge provided by process experts. This new methodology makes the contribution of offering a method to avoid the need to have a large training database during the development process for the diagnosticator, which will imply the use of a fault symptom table. In addition, the PMML standard is used to transmit the information of the model obtained in the configuration application to the application that performs the online monitoring tasks. The effectiveness of the methodology is analyzed in a real case, specifically in the diagnosis of failures of a water treatment process for injection in the pharmaceutical industry.

Keywords : fault diagnosis; decision trees; fuzzy logic; PMML standard; methodology proposal.

        · abstract in Spanish     · text in Spanish     · Spanish ( pdf )