Integration of Finite Element Method and Neural Network for Enhanced Prediction of Rubber Buffer Stiffness in Light Aircraft

Rubber buffers are one of the most important components for structural vibration damping in light aircraft. This study presents a finite element model developed using ABAQUS, which has been experimentally validated. The stiffness of rubber buffers with varying geometric parameters under different loading conditions was analyzed using ABAQUS. The stiffness of rubber buffers is predicted via a BP neural network model. A novel approach integrating the finite element method with neural network analysis is proposed. This method initially derives buffer stiffness data through the finite element model, which is subsequently utilized to train the neural network model for predicting rubber buffer stiffness. The results indicate that both geometric parameters and loading conditions significantly affect the stiffness of rubber buffers. The proposed integration of the finite element method and neural network analysis not only reduces time and economic costs but also enhances calculation accuracy, rendering it more suitable for engineering applications. Comparative analyses reveal that the prediction accuracy of the BP neural network ranges from 67.59% to 88.5%, which is higher than that of traditional formulas. Furthermore, the model demonstrates superior capability in addressing multivariate linear coupling relationships.