Dissertation > Excellent graduate degree dissertation topics show

Early evidence theory and its application in fault diagnosis information fusion research

Author: YangJianPing
Tutor: HuTianYou；HuangHongZhong
School: University of Electronic Science and Technology
Course: Mechanical Manufacturing and Automation
Keywords: Evidence Theory Conflict of evidence Information Fusion Early Failure
CLC: V263.6
Type: Master's thesis
Year: 2008
Downloads: 453
Quote: 7
Read: Download Dissertation

Abstract

Aviation aircraft engine as the most important parts, which directly affect the safety and reliability of the aircraft's flight performance, economy and safety of the crew. If the engine failure in the early stages of the detected out and dealt with, the engine will work effectively improve security and reliability. Therefore, it is necessary to expand early fault diagnosis of engine research. However, in the early stages of engine failure, the failure characteristics are very weak, and the emergence of fault characteristic uncertainty. In order to improve the accuracy of early diagnosis, a large number of real-time status monitoring engine resulting engine sensor data fusion of early failures to become a very necessary and effective passing. At present, with information technology, computer technology and sensor technology, DS evidence theory as an effective tool for uncertainty reasoning has been extensively studied and successfully applied to information fusion, which also makes the DS evidence theory based information fusion technology to become a very effective early aeroengine fault diagnosis approach. This paper from the National Natural Science Foundation Project: Based on the weak signal feature extraction and uncertain information fusion Early aeroengine fault diagnosis (item number: 60672165). Firstly, information fusion tool for DS evidence theory is studied. In the integration of evidence theory itself evidence of high conflict occurs when paradox arising from the integration of three aspects of the evidence for analysis: 1) evidence and Jiaoyuan Xin Ren assigned identical; 2) evidence and evidence of internal primitives are identical to assign trust different; 3) evidence of exactly the same and the internal evidence of primitive assign different trust. Evidence obtained from the three aspects of trust focused theoretical reasons. And on this basis, based on the idea of ??local conflicts partial allocation is proposed evidence-based differences in the absolute and relative factor primitives new paradox combination formula, and with the results of other synthetic methods were compared to results of a new approach focusing trust the effectiveness and robustness, effectively solve the evidence theory of evidence in the synthesis of high conflict occurs when paradox. Secondly, introduces the basic concepts of information fusion, levels, hierarchies, and the current information fusion method. Combined with the characteristics of fault diagnosis, fault diagnosis information fusion from the actual background and information theory point of view, illustrates the information fusion application in fault diagnosis solid mathematical foundation. Combined with the level of information fusion and the level of general troubleshooting information shows the basic structure and framework integration, and troubleshooting information fusion method are summarized. Finally, the DS evidence theory is applied to the early fusion aeroengine fault diagnosis is established early application of evidence theory for fault diagnosis model. This section focus on the following two aspects of the development: the theory of evidence applied to fault diagnostics feature layer information fusion, in the early fault diagnosis model BPA proposes a new construction method, obtained than a single sensor more accurate fault symptoms information; considered to be drawn in fault diagnosis fault symptom rather than the cause of the actual requirements, combined with the early stages of aeroengine fault characteristics, application fuzzy measure indicates the fault symptom. Finally, the application of fuzzy clustering analysis to the principle of maximum degree of membership criteria drawn aeroengine early failure.

Related Dissertations

CLC: > Aviation, aerospace > Aviation > Aeronautical Manufacturing Technology > Aero-engine manufacturing > Fault Analysis and
© 2012 www.DissertationTopic.Net Mobile