Akademik Veri Yönetim Sistemi
Materialpruefung/Materials Testing, cilt.67, sa.9, ss.1520-1527, 2025 (SCI-Expanded)
Optimization techniques are crucial in industrial engineering, particularly in addressing complex design and operational challenges. Traditional optimization methods often struggle with high computational costs, poor convergence rates, and multimodal fitness functions. To overcome these limitations, nature-inspired metaheuristic algorithms have gained popularity. This study introduces a modified artificial neural network-assisted superb fairy-wren optimization algorithm (MSFWOA) to enhance the search and exploitation capabilities of the standard superb fairy-wren optimizer. The algorithm integrates artificial neural networks (ANNs) to improve solution accuracy and convergence efficiency. The effectiveness of MSFWOA is demonstrated through its application to industrial optimization problems, including heat exchanger cost minimization, reinforced concrete beam structural optimization, piston lever volumetric optimization, pressure vessel design, and aircraft wing rib component structural optimization. Comparative analysis with existing metaheuristic algorithms highlights the superior performance of MSFWOA in achieving optimal solutions with reduced computational cost and higher precision.