Control Method of Main Steam Valve Opening in Multi-Machine Power System by Integrating Dynamic Surface and RBF Neural Network
This study proposes a robust control method for main valve opening in multi-machine power systems that integrates dynamic surfaces and radial basis function neural networks to address the poor robustness and stability in existing control systems. As a method of simplifying controller design by intro...
Saved in:
| Main Author: | |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2025-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/11036091/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | This study proposes a robust control method for main valve opening in multi-machine power systems that integrates dynamic surfaces and radial basis function neural networks to address the poor robustness and stability in existing control systems. As a method of simplifying controller design by introducing low-pass filters, dynamic surfaces can effectively reduce computational complexity. Meanwhile, radial basis function neural networks are used to handle unknown non-linear characteristics in the system, enhancing the adaptive ability and robustness of the controller. Due to the time delay in signal transmission during the operation of the power system, the dynamic performance may decrease. Therefore, a robust controller design method for the main steam valve opening based on the L<inline-formula> <tex-math notation="LaTeX">$\infty $ </tex-math></inline-formula> norm performance index is proposed. This method sets the initialization conditions for multi-machine power systems and combines dynamic surfaces and radial basis function neural networks to process functions with time delays in the system to meet the L<inline-formula> <tex-math notation="LaTeX">$\infty $ </tex-math></inline-formula> norm performance index of system tracking error and ensure system stability. The simulation results demonstrated significant improvements in system stability and robustness, with the rotor angle difference and angular velocity error converging quickly to stable values. After increasing the L<inline-formula> <tex-math notation="LaTeX">$\infty $ </tex-math></inline-formula> performance index, the controller still quickly stabilized the system under time delay conditions, and the rotor angle error and angular velocity error converged to 0.70° and 0rad/s, respectively, verifying its robustness performance. The combination of dynamic surface control and radial basis function neural network significantly improves the robustness of the system to time delay and interference. This method has important theoretical and practical application value in the stable operation of power systems. |
|---|---|
| ISSN: | 2169-3536 |