Optimization Of Fibre-Reinforced Composites Using Several Objectives: Machine Learning Approaches for Improved Structural Performance

Abstract

This paper explores fibre-reinforced composite material optimization with many objectives in a range of engineering uses. Because of their adjustable mechanical qualities and high strength-to-weight ratios, composite materials have gained widespread acceptance in the automobile, marine, aerospace, and defining sectors. The optimization of laminated composite structures presents unique challenges due to their anisotropic behaviour and numerous design variables, including fibre orientation, volume fraction, and layer sequence. While traditional approaches have focused on single-objective optimization, this research emphasizes the importance of multi-objective techniques that address the inherent trade-offs between competing factors such as weight reduction, cost efficiency, stiffness enhancement, and vibration control. Various optimization methods, including genetic algorithms, ant colony optimization, and machine learning approaches such as random forest regression, support vector regression, and ada boost regression, are explored for their effectiveness in navigating complex design spaces. The review highlights how proper optimization can significantly improve structural performance while maintaining damage tolerance and cost-effectiveness. The research contributes to the advancement of composite material design by demonstrating how systematic optimization across multiple parameters can yield significant performance improvements, addressing both mechanical requirements and stability goals in critical applications ranging from aircraft frames to wind turbine blades.