Multiple fault disambiguations through parameter estimation: a bond graph model-based approach
by S.K. Ghoshal, A.K. Samantaray
International Journal of Intelligent Systems Technologies and Applications (IJISTA), Vol. 5, No. 1/2, 2008

Abstract: To ensure safe operation of industrial processes, automated Fault Detection and Isolation (FDI) procedures are implemented in their supervision platforms. In the safety-critical and environmentally hazardous processes, it is impossible to introduce all kinds of faults and then to derive their consequences. Qualitative determination of consequences of different faults can be misleading in complex dynamical systems. Therefore, simulation of a prototype model turns out to be a practical and an economical solution for the development of a complete Knowledge-Base (KB). Consequently, the intelligence acquired by KB from the simulated models is used to fine-tune the Decision Support System (DSS) such that false alarms and misdetections are minimised. A method for model-based multiple FDI by using Analytical Redundancy Relations (ARRs) and parameter estimation is developed in this paper. Parameter estimation is an essential prerequisite for fault accommodation through system reconfiguration or Fault Tolerant Control (FTC). Bond graph modelling is used to describe the process models and then the model is used to derive the ARRs and fault candidates. Parameter values corresponding to the fault-subspace are estimated by minimising a function of the ARRs. Modelling uncertainties arising out of parameter estimation and sensor noise are taken care by using a passive approach for robust FDI. The developed technique is applied to monitor an open-loop non-linear thermo-fluid process.

Online publication date: Mon, 05-May-2008

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

 
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Intelligent Systems Technologies and Applications (IJISTA):
Login with your Inderscience username and password:

Username:        Password:          Forgotten your password?

Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email subs@inderscience.com