Multiple-scale uncertainty optimization design of hybrid composite structures based on neural network and genetic algorithm. (15th April 2021)

• Record Type: Journal Article
• Authors: Peng, Xiang; Qiu, Chan; Li, Jiquan; Wu, Huaping; Liu, Zhenyu; Jiang, Shaofei
• Is Part Of: Composite structures. Volume 262(2021)
• Publication Date: 2021-04-15
• Subjects: Hybrid composite structure -- Stacking sequence -- Material patch -- Neural network -- Genetic algorithm; Composite construction -- Periodicals; Composites -- Périodiques; 624.18
• DOI: 10.1016/j.compstruct.2020.113371
• Languages: English
• ISSNs: 0263-8223

Abstract

Abstract: A multiple-scale uncertainty optimization design framework is established for hybrid composite structures, where the stacking sequence and material patches are optimized simultaneously considering micro-scale uncertainty material and strength parameters. Firstly, the uncertainty propagation and performance calculation are implemented based on the finite-element method and classical laminated plate theory, and an adaptive neural network model is constructed to quantify the uncertainties of the natural frequency, reliability index and total cost. Furthermore, the uncertainty optimization design function for the hybrid composite structure is constructed, which maximizes the natural frequency under constraints of the reliability index and total cost, the stacking sequences and material patches are optimized using an adaptive genetic algorithm. In comparison with the determinate optimization results and optimum staking sequence with single materials in two examples with different in-plane loads and cost constraints, the proposed methodology improves the natural frequency by 2.14%–18.61% with the constraints of the reliability index and total cost under multiple scale uncertainties.