基于神经网络的弹性力学位移计算方法研究  

Research on Neural Network-Based Computational Methods for Elasticity Displacement in Mechanics

作　　者：蔡振荣 Cai Zhenrong(School of Architecture and Transportation Engineering,Guilin University of Electronic Technology,Guilin,China)

机构地区：[1]桂林电子科技大学建筑与交通工程学院,广西桂林

出　　处：《科学技术创新》2024年第6期104-107,共4页Scientific and Technological Innovation

摘　　要：常规的弹性力学问题是以有限元方法为主进行数值求解的,网格划分对计算结果影响较大,从而在求解力学问题时存在一定的局限性。神经网络作为一种通用数值逼近器,可以用来实现对弹性力学偏微分方程的求解。本文建立了基于物理约束的神经网络偏微分方程求解模型,以常见的二维平面压缩计算为例,实现了该方法的求解和验证,为智能化弹性力学求解提供了新思路与方法。Conventional elasticity problems are primarily numerically solved using finite element methods,where mesh partitioning significantly influences computational results,leading to certain limitations in solving mechanical problems.Neural networks,as a versatile numerical approximator,can be employed to solve partial differential equations in elasticity mechanics.This paper establishes a neural network partial differential equation solving model based on physical constraints.Taking the example of common two-dimensional plane compression calculations,the method's solution and validation are implemented,providing new ideas and methods for intelligent elasticity problem-solving.

关 键 词：弹性力学 神经网络 数值计算 偏微分方程组 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]