A synergy-effect-incorporated fuzzy Petri net modeling paradigm with application in risk assessment

Xiaoliang Wang, Faming Lu, Meng Chu Zhou, Qingtian Zeng

Research output: Contribution to journalArticlepeer-review

15 Scopus citations

Abstract

As a graphical knowledge representation and reasoning tool, fuzzy Petri net (FPN) is widely used in risk assessment, fault diagnosis and many other fields. It can express the complex causal relationships among events and make fuzzy reasoning about them. In practical applications, there may be synergy effects among events, i.e., the occurrence of event a may increase or decrease the truth degree of event b. Traditional FPNs fail to express such effects, thus neglecting them during their reasoning procedure. To address this issue, this study proposes an FPN considering them for the first time. Firstly, this study illustrates the concept of synergy effects via practical application examples. Secondly, it develops the transition enabling conditions and truth degree calculation rules incorporated with synergy effects. Then, it formally defines a synergy-effect-incorporated fuzzy Petri net (SFPN). On this basis, it presents the production rules, transition firing rules and reasoning algorithm of SFPN to perform knowledge representation and reasoning for knowledge-based systems affected by synergy effects. Finally, it takes the risk assessment of equipment failure caused by corrosion as a running case to validate its feasibility and advantages.

Original languageEnglish (US)
Article number117037
JournalExpert Systems with Applications
Volume199
DOIs
StatePublished - Aug 1 2022

All Science Journal Classification (ASJC) codes

  • General Engineering
  • Computer Science Applications
  • Artificial Intelligence

Keywords

  • Expert system
  • Fuzzy Petri net
  • Fuzzy reasoning
  • Risk assessment
  • Synergy effect

Fingerprint

Dive into the research topics of 'A synergy-effect-incorporated fuzzy Petri net modeling paradigm with application in risk assessment'. Together they form a unique fingerprint.

Cite this