Dynamic Compensation and fault Detection and Isolation in Thermocouples
N. Divya1, Y. Vidyasagar2

1N. Divya, Student, Anna University CEG, Chennai (Tamil Nadu), India.
2Y. Vidyasagar, Student, SRM University Nagari, (Andhra Pradesh), India
Manuscript received on 12 May 2019 | Revised Manuscript received on 06 June 2019 | Manuscript Published on 15 June 2019 | PP: 232-236 | Volume-8 Issue-1S3 June 2019 | Retrieval Number: A10410681S319/2019©BEIESP
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Abstract: For the safe and efficient operation of the many industries where temperature measurement is required such as power generation, cement etc., it is important to use temperature as the feedback signal in the control systems. As temperature measurement is done with thermocouples in most of the industries including power generation, Response time of thermocouple has to be measured periodically. In order to avoid the human intervention, and the errors associated with it, It is always preferred to be done online i.e., while the plant is in operation. To suggest a solution to this problem a scheme aiming at the Dynamic compensation of thermocouple in order to improve time response characteristics and to reduce the noise in the output signal and also fault detection and isolation of thermocouple to handle the degradations is presented in this paper. In this scheme as a part of the dynamic compensation, the Kalman filter formulation in linear time-invariant system framework for the thermocouples is used to assess and regenerate the input temperature to a thermocouple. The model of K and E type thermocouples are used for the demonstration of the results and by using these models the filter parameters are tuned appropriately to achieve the desired response. Generalized likelihood ratio (GLR) method is used in the Fault detection and isolation algorithm to identify the magnitude of fault in the thermocouples. The correction in the thermocouple output to compensate for the slow response and unwanted abrupt jumps is found to be extremely satisfactory.
Keywords: Dynamic Compensation, Fault Detection and Isolation, Kalman Filter, Response Time, Thermocouple.
Scope of the Article: Foundations Dynamics