Research On Coupled Multiple-Fault Diagnosis Based On Dezert-Smarandache Theory

With the rapid development of mass production and complexity in process industries, reliability and security are being greatly needed to avoid large economical loss brought by accidents and even breakdowns of industrial productions. In complex industrial process (such as rolling process), multiple faults may occur simultaneously, among which, there are always notorious characteristics, such as coupling and transfer, etc. For the complex industrial process, the research of multiple-fault diagnosis methods not only has important academic value, but also meets the demand of complex industrial process production and operation, so it has very important significance to monitor and ensure the safety in production and product quality. It has been proved that Dempster-Shafer Theory (DST) has been widely applied in single fault diagnosis problems and Dezert-Smarandache Theory (DSmT) is a new theory based on it. DSmT provides a unified frame of discernment for single fault and simultaneous faults, representing, measuring and fusing the uncertainty.In this thesis, based on the DS theory and related to steel rolling process as the research background, through extensive review of related literature, the methods of coupling fault diagnosis are presented and studied systematically, and the main research work is as follows:Firstly, define the multiple-fault and the characteristics of it. The basic principle of evidence theory is introduced in detail, and fuzzy membership degree is applied to obtain evidence. Research on the specific methods of coupling multiple fault diagnosis algorithms based on the theory of the classical DSmT, using the test data for simulation, according to the rules to determine the fault modes.Secondly, for the computing and storage problems of classical DSmT, mixed DSmT is studied in theory. Adding a priori knowledge on the basis of the classical theory of DSm makes the mixed DSmT more actual, and at the same time be able to deal with the evidence with larger uncertainty. Through the test data are simulated and compared with the classical theory of DSm analysis, the research verifies the superiority of mixed DSmT, which is then applied to the coupling multiple-fault diagnosis of rolling process.Finally, the priori knowledge is limited and leads to incomplete identification framework, combination rules of open identification framework are put forward. The two-dimensional conflict measure factors of DST are extended to DSmT framework, and use that to determine to choose the combination rules of closed or open framework; and then do the diagnosis experiment with the field data of rolling process. The experimental results show that the proposed method can effectively solve the conflict evidence due to incomplete framework. Therefore it has practical application value.