Robust Load Frequency Control of Interconnected Power Systems with Back Propagation Neural Network-Proportional-Integral-Derivative-Controlled Wind Power Integration

Fang Ye, Zhijian HuSustainability 2024 · DOI

TL;DR

This paper tackles the challenge of grid frequency instability caused by the increasing integration of wind power. It introduces a novel two-pronged approach: a BPNN-PID controller to smooth wind power output and an H∞ robust controller to maintain grid frequency stability in a multi-area power system. The proposed methods effectively mitigate the impact of wind power fluctuations, ensuring a more reliable and stable power grid.

What Problem Does This Paper Address?

The global push for sustainable energy sources has led to a significant increase in wind power integration. However, wind power is inherently intermittent and fluctuates due to unpredictable wind speeds. These fluctuations can destabilize the power grid's frequency, posing a significant challenge to maintaining a reliable electricity supply. Traditional Load Frequency Control (LFC) methods struggle to cope with the magnitude and speed of these wind-induced disturbances, especially in complex, interconnected multi-area power systems (MAPS). This paper aims to address this gap by developing advanced control strategies that can effectively manage wind power variability and ensure grid frequency stability.

Methods

The paper proposes a two-stage control strategy:

  1. Wind Power Output Smoothing (BPNN-PID Controller): To address the volatility of wind power output, a Back Propagation Neural Network-Proportional-Integral-Derivative (BPNN-PID) controller is introduced. This controller uses a three-layer neural network to adaptively tune the PID controller's gains in real-time. The BPNN learns from the system's output and error signals, allowing it to predict and adjust the wind turbine's output to better match the desired rated power, thereby reducing fluctuations.
  2. Grid Frequency Stability (H∞ Robust Controller): To mitigate the impact of any remaining imbalanced wind power on the main grid, an H∞ robust controller is designed. This controller operates at the multi-area power system level and is formulated using linear matrix inequalities (LMIs). The H∞ control framework is chosen for its ability to guarantee a certain level of performance (specifically, bounding the frequency deviation) even in the presence of uncertainties and disturbances, such as those introduced by wind power.

Datasets

Not discussed in the paper. The paper uses a simulated four-area interconnected power system for validation.

Key Insights and Ideas

  • Adaptive PID Tuning is Crucial: Traditional PID controllers, while widely used, are often manually tuned and lack the adaptability needed for rapidly changing renewable energy sources. The BPNN-PID approach demonstrates that adaptive gain tuning is essential for effectively managing wind power fluctuations.
  • Two-Layered Control for Robustness: Combining a local controller (BPNN-PID) for smoothing individual wind turbine output with a global controller (H∞ robust controller) for overall grid stability provides a comprehensive solution. This layered approach addresses both the source of the disturbance and its impact on the larger system.
  • H∞ Control for Guaranteed Performance: The H∞ robust controller provides a theoretical guarantee on the system's frequency deviation, ensuring it stays within acceptable limits despite wind power uncertainties. This is a significant advantage over controllers that might only offer performance under nominal conditions.
  • BPNN-PID's Faster Response: The BPNN-PID controller exhibits a faster response in tracking desired wind power output compared to conventional PID, which is critical for quickly counteracting wind speed changes.

Key Results

  • The BPNN-PID controller demonstrated a faster response in tracking rated wind power compared to a conventional PID controller.
  • The proposed robust controller significantly improved the frequency response of the four-area interconnected power system. In the closed-loop system, regional frequency deviations converged rapidly, unlike the diverging behavior observed in the open-loop system.
  • The robust controller outperformed a traditional LFC controller in terms of regional frequency error response and inter-area power exchange.
  • The study validated the feasibility and effectiveness of both the BPNN-PID and the H∞ robust controllers in mitigating the impact of wind power integration on grid frequency stability, meeting the H∞ performance criteria.

Limitations and Future Work

  • Simulation-Based Validation: The study relies on simulations of a four-area power system. While the results are promising, real-world implementation and testing on actual power grids would be the next logical step.
  • Focus on Wind Power: The current work primarily focuses on wind power integration. Future research could explore extending these control strategies to other variable renewable energy sources like solar power.
  • Machine Learning for Sustainable Energy: The authors express interest in further exploring machine learning techniques for sustainable energy control to enhance multi-regional frequency stability. This could involve investigating more advanced AI algorithms or hybrid approaches.

Summary updated

Abstract

As the global demand for energy sustainability increases, the scale of wind power integration steadily increases, so the system frequency suffers significant challenges due to the huge fluctuations of the wind power output. To address this issue, this paper proposes a Back Propagation Neural Network-Proportional-Integral-Derivative (BPNN-PID) controller to track the output power of the wind power generation system, which can well alleviate the volatility of the wind power output, resulting in the slighter imbalance with the rated wind power output. Furthermore, at the multi-area power system level, to mitigate the impact of the imbalanced wind power injected into the main grid, the H∞ robust controller was designed to ensure the frequency deviation within the admissible range. Finally, a four-area interconnected power system was employed as the test system, and the results validated the feasibility and effectiveness of both the proposed BPNN-PID controller and the robust controller.
BibTeX
@inproceedings{ye2024robust,
  author = {Fang Ye and Zhijian Hu},
  title = {{Robust Load Frequency Control of Interconnected Power Systems with Back Propagation Neural Network-Proportional-Integral-Derivative-Controlled Wind Power Integration}},
  booktitle = {Sustainability},
  year = {2024},
  doi = {10.3390/su16188062},
}
Figure