Neuro-Fuzzy Technique Based Fault Diagnosis in Shell and Tube Heat Exchanger

Abstract

In order to transfer heat between two different medium or else the same medium heat exchanger is used. It is most widely used in oil refineries and other large chemical processes, and it is also suitable for high-pressure applications. Because of the change in process parameters, the general warmth exchange rate will be reduced in the heat exchanger. In order to rectify it, the possible faults developed in the heat exchanger have to be identified. Due to this reason, various fault diagnosis methods are commonly used. An Adaptive Neuro-Fuzzy Inference System (ANFIS) is preferred here. This method requires large amount of process data. The real time readings for both normal and fault operating conditions from a shell and tube heat exchanger are obtained. In addition, a transfer function model is also derived using the real time data. The ANFIS is trained with both the normal and fault data. By entering the test input data, the working condition of the heat exchanger can be identified by this technique.

Keywords: fault diagnosis, shell and tube heat exchanger, Adaptive Neuro-Fuzzy Inference System

Cite this Article

Monisa S, Vijayachithra S. Neuro-Fuzzy Technique Based Fault Diagnosis in Shell and Tube Heat Exchanger. Journal of Instrumentation Technology and Innovations. 2016; 6(2): 21–27p.