D-S Evidence Theory and Its Application for Fault Diagnosis of Machinery

With the development of science and technology, gas turbines have brought huge economic benefits as modern power equipment. The gas turbine fault prognostics health management (PHM) technology is an important foundation to ensure its safe and stable operation. In view of the current insufficient utilization of multi-source monitoring data for gas turbines, the difficulty of eliminating sensor signal uncertainty, and the need to improve diagnosis accuracy, Dempster-Shafer (D-S) evidence theory is a concise and efficient decision-level multi-source in-formation fusion method. It has huge application potential in gas turbine condition monitoring and fault diagnosis. This article introduces the traditional D-S evidence theory and, in view of its shortcomings, combs the classic improvement methods of evidence theory. Then the research status of the evidence theory in the field of gas turbine fault diagnosis is systematically summarized. Finally, this article focuses on the actual industrial application scenarios of gas turbines, analyzes, and summarizes the characteristics of D-S evidence theory, and points out the future development trend of D-S evidence theory and gas turbine fault diagnosis.