Boiler modeling and controller performance optimization using genetic algorithm for palm oil industry
This research developed control algorithms and Graphical User Interfaces (GUI) using Genetic Algorithm (GA) optimization analysis for the boiler control system. The trade-off optimized PI controller tunings visualized by Adaptive GA Optimization Contol Toolboxes provided the best control performance...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English English |
| Published: |
2020
|
| Subjects: | |
| Online Access: | https://eprints.ums.edu.my/id/eprint/40726/1/24%20PAGES.pdf https://eprints.ums.edu.my/id/eprint/40726/2/FULLTEXT.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | This research developed control algorithms and Graphical User Interfaces (GUI) using Genetic Algorithm (GA) optimization analysis for the boiler control system. The trade-off optimized PI controller tunings visualized by Adaptive GA Optimization Contol Toolboxes provided the best control performance in terms of settling time and integral error values for both servo and regulatory controls. Routh-Hurwitz necessity criterion analysis has been applied to determine the stability margins. This criterion has restricted the search region of GA to ensure proper searching of chromosomes to minimize the period for optimization analysis and avoiding the optimum PI tunings values from missing out on any search region. The conducted simulation and validation tests have shown that Adaptive GA Optimization Analysis Toolboxes has provided better PI tunings to improve the control performance indexes up to 84.61% for the simulation analysis and 93.25% improvement on the validation tests. In addition, the settling time of the control responses have improved up to 80.18% for simulation analysis and 83.49% for validation test. The reason is due to Adaptive GA Optimization has applied stochastic optimization technique, which is repetitively proposed individuals or chromosomes to be tested using objective function in the computation analysis and then, will choose the controller tunings with least integral error values for both servo and regulatory controls. It offers better tuning opportunity without relying on the fixed tuning formulas as performed by manually calculated controller tunings. |
|---|