Scholarlink

This study examines the bending behaviour of functionally graded material (FGM) beams using a real physical discrete model. The beams are subjected to point loads applied at different positions. The discrete model is utilised to generate a comprehensive dataset of displacements under various loading conditions. This dataset is then employed to train an artificial neural network (ANN) using the Adam optimiser to predict beam displacements with high accuracy. The proposed approach harnesses the efficiency of ANN-based modelling to deliver fast and reliable displacement predictions, presenting a robust alternative to conventional numerical methods. The findings highlight the ANN’s capability in capturing the complex bending response of FGM beams, demonstrating its potential for advanced structural analysis applications.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

ANN with Adam Optimizer Combined with a Discrete Model for Analysing the Static Bending of FGM Beams

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Ihsan Tikonab,
  • Anass Moukhliss,
  • Abdellatif Rahmouni,
  • Rhali Benamar

摘要

This study examines the bending behaviour of functionally graded material (FGM) beams using a real physical discrete model. The beams are subjected to point loads applied at different positions. The discrete model is utilised to generate a comprehensive dataset of displacements under various loading conditions. This dataset is then employed to train an artificial neural network (ANN) using the Adam optimiser to predict beam displacements with high accuracy. The proposed approach harnesses the efficiency of ANN-based modelling to deliver fast and reliable displacement predictions, presenting a robust alternative to conventional numerical methods. The findings highlight the ANN’s capability in capturing the complex bending response of FGM beams, demonstrating its potential for advanced structural analysis applications.