Online fault diagnosis for smart machines embedded in Industry 4.0 manufacturing systems: A labeled Petri net-based approach. (June 2021)

• Record Type: Journal Article
• Authors: Paiva, Pedro R.R.; de Freitas, Braian I.; Carvalho, Lilian K.; Basilio, João C.
• Is Part Of: IFAC journal of systems and control. Volume 16(2021)
• Publication Date: 2021-06
• Subjects: Online diagnosis -- Petri net models -- Industry 4.0 -- Smart manufacturing systems; Automatic control -- Periodicals; Relay control systems -- Periodicals; Embedded computer systems -- Periodicals; Feedback control systems -- Periodicals; Artificial intelligence -- Periodicals; Artificial intelligence; Automatic control; Embedded computer systems; Feedback control systems; Relay control systems; Electronic journals; Periodicals; 629.89
• DOI: 10.1016/j.ifacsc.2021.100146
• Languages: English
• ISSNs: 2468-6018

Abstract

Abstract: Detection of abnormality (or faults) occurrences is of paramount importance in smart manufacturing systems of Industry 4.0 since faults do not usually take the system immediately to a halt, and so, it can jeopardize an entire production. With that in mind, we propose here an online diagnoser based on the Petri model of either a specific machine or part of a smart manufacturing system that makes its decision regarding the fault occurrence by storing the sequence of observed events and, after each new occurrence of an observable event, it updates its state by verifying if two sets of inequalities are satisfied: one set that accounts for the normal behavior, and another one for the faulty behavior. The main advantage of the method proposed here over existing ones are as follow: (i) it requires simple inequality verification, as opposed to online solution of Integer Linear Programming Problems; (ii) it allows different transitions to be labeled by the same event, as opposed to one-to-one event-transition labeling previously assumed, which is a serious limitation, as far as smart manufacturing systems is concerned. The effectiveness of the proposed method is illustrated by applying it to a hypothetical machine embedded in a smart manufacturing line and comparing its performance with a method previously proposed in the literature.