Training of Feed-Forward Neural Networks by Using Optimization Algorithms Based on Swarm-Intelligent for Maximum...

Area of Science:

• Artificial Intelligence
• Renewable Energy Systems
• Optimization Algorithms

Background:

• Artificial neural networks (ANNs) are crucial for maximum power point tracking (MPPT) in renewable energy.
• Effective MPPT relies heavily on the successful training of ANNs.
• Metaheuristic algorithms, particularly swarm intelligence, are widely used for ANN training.

Purpose of the Study:

• To evaluate and rank 13 swarm-intelligent optimization algorithms for training feed-forward neural networks.
• To determine the most effective algorithms for achieving accurate maximum power point tracking.
• To assess algorithm performance across different neural network structures using mean squared error.

Main Methods:

• Utilized 13 swarm-intelligent algorithms: artificial bee colony, butterfly optimization, cuckoo search, chicken swarm optimization, dragonfly algorithm, firefly algorithm, grasshopper optimization algorithm, krill herd algorithm, particle swarm optimization, salp swarm algorithm, selfish herd optimizer, tunicate swarm algorithm, and tuna swarm optimization.
• Trained feed-forward neural networks for MPPT using these algorithms.
• Evaluated performance based on mean squared error (MSE) during training and testing phases.

Main Results:

• The firefly algorithm, selfish herd optimizer, and grasshopper optimization algorithm were ranked as the top three most successful algorithms.
• Achieved low training and testing mean squared errors: 4.5 × 10^-4, 1.6 × 10^-3, and 2.3 × 10^-3 for training, and 4.6 × 10^-4, 1.6 × 10^-3, and 2.4 × 10^-3 for testing, respectively.
• Effective results were obtained with a low number of evaluations, indicating efficiency.

Conclusions:

• The firefly algorithm, selfish herd optimizer, and grasshopper optimization algorithm are highly effective for ANN training in MPPT.
• The evaluated swarm-intelligent algorithms generally provide acceptable results for MPPT applications.
• These algorithms demonstrate significant potential for improving MPPT efficiency in renewable energy systems.