An optimal neural network to design generators and stabilizers for multi-machine power systems based on a promoted...

Area of Science:

• Electrical Engineering
• Power Systems Analysis
• Control Systems

Background:

• Power system stabilizing (PSS) is crucial for maintaining stability in multi-machine power systems.
• Generator and network modeling significantly impact PSS design and performance.
• The presence or absence of an infinite bus affects system dynamics and PSS effectiveness.

Purpose of the Study:

• To investigate the role of generator and network modeling in PSS design for multi-machine power systems.
• To develop and validate an optimized artificial neural network (ANN) based PSS using a promoted firefly algorithm (PFF).
• To assess the performance of the proposed ANN/PFF-PSS in damping oscillations and improving voltage recovery under various operating conditions.

Main Methods:

• Simulation of different generator and network models, including systems with and without an infinite bus.
• Utilization of an optimal artificial neural network (ANN) where PID controller parameters are the network output.
• Optimization of the ANN using a promoted firefly algorithm (PFF) for PSS design in multi-machine power systems.

Main Results:

• The proposed ANN/PFF-PSS significantly reduces load angle overshoot by 35.7% and settling time by 28.6% compared to conventional PSS.
• Voltage recovery is improved by 9.3% with the ANN/PFF-PSS.
• Robustness analysis confirms the effectiveness of the proposed stabilizer in damping both inter-area and intra-area oscillations.

Conclusions:

• The ANN/PFF-PSS demonstrates superior performance in enhancing dynamic stability for multi-machine power systems.
• The study validates the importance of accurate modeling and the effectiveness of advanced optimization techniques for PSS design.
• The proposed method offers a robust solution for damping oscillations in complex power networks.