Dissertation > Excellent graduate degree dissertation topics show

Fault Diagnosis for Linear Time Delay System Based on Observer Approach

Author: YangXiaoBo
Tutor: ShiHongRui
School: Donghua University
Course: Detection Technology and Automation
Keywords: Fault Diagnosis Delay systems Observer Residual system
CLC: TP18
Type: Master's thesis
Year: 2009
Downloads: 79
Quote: 0
Read: Download Dissertation

Abstract

With the ever-increasing demands of the dynamic system security and reliability, and fault diagnosis in recent years received more attention, including the model-based approach is a hot research field. However, most of the achievements in the field of fault diagnosis study for non-delay systems. Delay phenomenon is inherent in the control system a feature widely in a variety of actual systems, such as the chemical production process, the papermaking industry, based on a control system of the network, traffic control, etc.. In addition, many large time constant of the system, also used appropriate time constant and pure lags to approximate, which can be attributed to the delay system model. The existence of time delays makes system analysis and synthesis of more complex and difficult at the same time, the time delays are often the root causes system instability and poor system performance. So is particularly important for the study of the delayed fault diagnosis method. In this paper, the development of fault diagnosis techniques do more detailed presentation summarizes the progress and results of the fault diagnosis method, pointed out that the delayed fault detection system to be perfect: the uncertainty of robustness and the sensitivity of the fault. Proposed fault diagnosis method based on unknown input observer for a class with the failed state linear systems. Time delay or interference terms decoupling and non-decoupling observer gain matrix and the residual system design, and then select the appropriate threshold for fault diagnosis. Robust fault diagnosis observer design problem for linear time-delay system modeling errors and unknown input disturbances exist, into the design of robust fault diagnosis observer model matching problem, arising from the residuals of the model uncertainties and unknown input disturbance to maintain the robustness, treat the failure to have a higher sensitivity of detection can achieve better diagnostic results. H_ ∞ performance index, the application of linear matrix inequalities (LMI) technology, is derived and proved sufficient condition for the existence of robust fault diagnosis observer design formulas. Interference and nonlinear perturbation containing unknown inputs for a class of delay systems to study adaptive observer-based fault diagnosis system, given in the form of linear matrix inequalities conditions for the existence of solutions of this problem and algorithm, and argues robust stability of the system and fault detection with high sensitivity and false positive rate of zero threshold ε judgment formula. The design of the adaptive observer having a dual function: first as fault detection observer is able to detect the occurrence of the fault, and then as a fault diagnosis observer can ideally estimated fault the shape changes with time. Man done a lot of simulation studies, the simulation results to verify the effectiveness of the proposed algorithm. Finally, the subject summarize and Next topic research directions prospect.

Related Dissertations

CLC: > Industrial Technology > Automation technology,computer technology > Automated basic theory > Artificial intelligence theory
© 2012 www.DissertationTopic.Net Mobile