EMPIRICAL MODELLING AND FUZZY CONTROL SIMULATION OF A HEAT EXCHANGER

• Main Author: OTHMAN, MOHD ISA
• Format: Final Year Project
• Language: English
• Published: Universiti Teknologi Petronas 2004
• Subjects: TK Electrical engineering. Electronics Nuclear engineering
• Online Access: http://utpedia.utp.edu.my/7186/1/2004%20-%20EMPIRICAL%20MODELLING%20AND%20FUZZY%20CONTROL%20SIMULATION%20OF%20A%20HEAT%20EXCHANGER.pdf; http://utpedia.utp.edu.my/7186/
• Institution: Universiti Teknologi Petronas

Aheat exchanger is one of the most important systems that have been installed in many process plants. It is a device that transfers heat from liquid to another without allowing them to mix. In order to ensure its smooth operation, modelling and simulation can be made so that its performance can beanalyzed and improved. At Process Control Lab, there is no simulation model for laboratory-scale heat exchanger pilot plant. Most ofthe time, the plant is being used for ordinary laboratory practice and the performance of this plant is not being analyzed. This project is therefore conducted to study the plant behavior and to optimize its performance by simulating it withnewtype of controller. The first goal of this project is to model the heat exchanger pilot plant by using empirical modelling method. It will yield the plant transfer function, GP that can be used for temperature controller analysis. Besides empirical modelling, mathematical modelling is also being carried out to study the heat exchanger behavior. By having the model, there is an alternative way to obtain forecasted data and result without extra cost. The second part of this project is to analyze the model temperature controller performance. Two controllers are being compared, namely PID and Fuzzy Logic Controller. First, PID controller is tested to yield the best tuning parameters for control valve. Ziegler-Nichols and fine tuning method is used to serve this purpose. Next, the data from PI controller simulation is fed into ANFIS toolbox in MATLAB for adaptive learning process. The FIS generated by ANFIS is based on Takagi- Sugeno fuzzy model. The FIS which is subsequently used by the Fuzzy Logic Controller will imitate the PI controller performance and perform based on range of data it has been trained before by ANFIS toolbox. Finally, the comparison between both controllers is concluded where Fuzzy Logic Controller is successfully imitating the PI controller with slightly better performance in terms of rise time, settling time and overshoot percentage.