Scholarlink

Modern combat modeling is accompanied by a large number of factors that make such models nonlinear. In particular, this is due to the maneuverability of troops and their interaction with the terrain, the limited effectiveness of attacks in large armies, reinforcements, logistics, etc. This paper presents machine learning methods for approximate solution of such conflicts under of conditions of non-linearity, considering resources models using the reaction-diffusion equations. This approach solves the problem of finding possible solutions to such equations. The proposed method uses physically-informed neural networks (PINNs) to approximate generalized solutions.

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

Deep Learning-Based Approximation of Nonlinear Military Conflict Models via Reaction-Diffusion Equations

  • 章节

  • 发表日期：

    2026年

查看全文
  • Michael Z. Zgurovsky,
  • Pavlo O. Kasyanov,
  • Liudmyla B. Levenchuk,
  • Vladyslav R. Novykov,
  • Liliia S. Paliichuk

摘要

Modern combat modeling is accompanied by a large number of factors that make such models nonlinear. In particular, this is due to the maneuverability of troops and their interaction with the terrain, the limited effectiveness of attacks in large armies, reinforcements, logistics, etc. This paper presents machine learning methods for approximate solution of such conflicts under of conditions of non-linearity, considering resources models using the reaction-diffusion equations. This approach solves the problem of finding possible solutions to such equations. The proposed method uses physically-informed neural networks (PINNs) to approximate generalized solutions.