Fuzzy logic controller for glycerin bleaching temperature control / Zakariah Yusuf
Bleaching process is one of the vital parts in the glycerin purification process to remove the color from the crude glycerin. Glycerin needs proper bleaching temperature control to avoid quality degradation of the produced oil. The temperature setting must be adequate to maximize the absorption but...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2012
|
| Online Access: | https://ir.uitm.edu.my/id/eprint/79011/1/79011.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Bleaching process is one of the vital parts in the glycerin purification process to remove the color from the crude glycerin. Glycerin needs proper bleaching temperature control to avoid quality degradation of the produced oil. The temperature setting must be adequate to maximize the absorption but it must not exceed the optimal point. The conventional method of controlling this kind of process is incapable of giving satisfactory result. This thesis presents two Fuzzy Logic Controllers, which are Fuzzy PD and Fuzzy PI to accommodate the problem. The controllers were designed with different membership functions and rules in order to achieve the desired performance. Furthermore, it will be integrated with Smith predictor. The performance of the proposed fuzzy logic controllers are benchmarked against PID in term of rise time, percentage overshoot and settling time. Error criterion evaluation such as IAE, ISE and ITAE are used to evaluate the controller performance. Self-tuning Fuzzy Logic controller is proposed to counter the mismatch model for Smith predictor. The simulation results indicate that fuzzy controllers gave better control performance compared to PID. Fuzzy PD controllers provide excellent performance with fast settling time with no overshoot were recorded. The application of Smith predictor shows improvement in the step response performance and error criteria. Self-tuning fuzzy logic controller managed to reduce the mismatch delay problem for Smith predictor structure. Self-tuning Fuzzy PD controller can track the set point up to 100% mismatch delay from the process model while self-tuning Fuzzy PI controller is only capable to track up to 50% mismatch delay. |
|---|