Fractional Order PID Controller Based‐Neural Network Algorithm for LFC in Multi‐Area Power Systems

• Published in: Engineering reports (Hoboken, N.J.) Vol. 7; no. 2
• Main Authors: El‐Rifaie, Ali M., Abid, Slim, Ginidi, Ahmed R., Shaheen, Abdullah M.
• Format: Journal Article
• Language: English
• Published: Hoboken, USA John Wiley & Sons, Inc 01.02.2025; Wiley
• Subjects: Algorithms; Control systems design; Controllers; Dynamic loads; Evolutionary computation; fractional‐order PID controller; Frequency control; Frequency stability; Genetic algorithms; grid stability; Hydrologic cycle; Load fluctuation; multi‐area power systems; Neighborhoods; neural network algorithm; Neural networks; Optimization; optimization techniques; Power plants; Proportional integral derivative; Systems stability
• ISSN: 2577-8196; 2577-8196
• DOI: 10.1002/eng2.70028

ABSTRACT Modern power systems are increasingly challenged by frequency stability issues due to dynamic load variations and the growing complexity of interconnected networks. Traditional PID controllers, while widely utilized, struggle to address the rapid fluctuations and uncertainties inherent in contemporary multi‐area interconnected power systems (MAIPS). This paper introduces an innovative approach to Load Frequency Control (LFC) using a Fractional‐Order PID (FOPID) controller, optimized by a Neural Network Algorithm (NNA). The proposed NNA‐FOPID framework leverages the biological principles of neural networks to dynamically tune controller parameters, significantly enhancing system performance. The solution is tested under various scenarios involving step load changes across multi‐area systems. The proposed method demonstrates marked improvements over traditional PID controllers and advanced optimization techniques such as Differential Evolution (DE) and Artificial Rabbits Algorithm (ARA). The comparisons show that the FOPID controller's NNA‐based design effectively and successfully handles LFC in MAIPSs for ITAE minimizations, and statistical evaluation supports its superiority. Motivated by artificial neural networks, this paper develops the NNA to optimize a Fractional‐Order PID controller for LFC in multi‐area interconnected power systems. It improves frequency regulating capabilities and reduces overshoots/undershoots. It achieves a 27.66%–93.78% improvement in Integral of Time Absolute Error compared to other techniques.