Load Frequency Control Optimization for Hydroelectric Systems Based on Type-II Fuzzy Deep Learning
This paper will propose a novel technique for optimizing hydroelectric (hydropower plants) on a small scale based on load frequency control that uses the self-tuning fuzzy proportional derivative method. Due to the fact that frequency is not controlled by any dump load or something else, this power...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Bilijipub publisher
2023-12-01
|
| Series: | Journal of Artificial Intelligence and System Modelling |
| Subjects: | |
| Online Access: | https://jaism.bilijipub.com/article_186539_08e928c5f35fc5ac91494d23e0d903d7.pdf |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | This paper will propose a novel technique for optimizing hydroelectric (hydropower plants) on a small scale based on load frequency control that uses the self-tuning fuzzy proportional derivative method. Due to the fact that frequency is not controlled by any dump load or something else, this power plant is under dynamic frequency variations that will use proportional-derivative controllers, which are optimized by fuzzy rules and then with deep learning techniques. The main purpose of this work is to maintain frequency in small hydropower plants at a nominal value. So, the proposed controller means Fuzzy Type-II-Proportional-Derivative will be used for load frequency control in a small hydropower system. The proposed schema can be used in different designations for both diesel generators and mini-hydropower systems at low stream flow. It is also possible to use a diesel generator in the hydropower system, which can be turned off when consumer demand is higher than electricity generation. The simulation will be done in MATLAB/Simulink to represent and evaluate the performance of this control schema under dynamic frequency variations. Spiking Neural Networks are used as the main deep learning techniques for optimizing this load frequency control, which turns into a Deep Spiking Neural Network. The obtained results indicated that the proposed schema has robust and high-performance frequency control in comparison to other methods. |
|---|---|
| ISSN: | 3041-850X |