摘要
Abstract
Additive manufacturing(AM)has demonstrated substantial application potential in aerospace,automotive,bio-medical,and high-end equipment industries,owing to its exceptional geometric freedom and customization capabilities.However,for metal additive manufacturing,the efficient design of high-performance structures that satisfy specific func-tional requirements,along with the accurate prediction of their mechanical behavior and service performance under complex working conditions,remains a critical bottleneck hindering its widespread industrial adoption.This paper presents a concise review of recent research on artificial intelligence surrogate models,constrained by both data-driven approaches and physi-cal mechanisms,in laser-based metal additive manufacturing.The review primarily covers typical applications in structural design optimization,rapid prediction of macroscopic properties,and correlation analysis among process parameters,micro-structure,and macroscopic properties.Furthermore,several recommendations are provided regarding the current challenges and corresponding strategies in the integration of laser metal additive manufacturing with artificial intelligence.The objec-tive is to offer valuable insights and references for researchers and engineers in related fields,thereby promoting the develop-ment of metal additive manufacturing toward greater intelligence,efficiency,and reliability.关键词
金属增材制造/人工智能代理模型/机器学习/深度学习/物理信息神经网络Key words
metal additive manufacturing/artificial-intelligence surrogate models/machine learning/deep learning/physics-informed neural networks分类
机械制造
