Scholarlink

<p>In order to study the Anti-Penetration Randomness of Metal Protective Structures (APRMPS) for the penetration probabilities of Metal Protective Structures under the action of the basic random variables, this paper analyzes the candidates for the basic random variables and the random response of APRMPS, and, on the basis of the improvement of Genetic Algorithm, proposes Dynamic Lifecycle Genetic Algorithm, including its main processes of the optimization of Back Propagation Neural Network. And by adopting the Back Propagation Neural Network optimized by Dynamic Lifecycle Genetic Algorithm (DLGABPNN) as the surrogate model of APRMPS, this paper presents the technical route of DLGABPNN-MCS, the Monte Carlo Simulation with DLGABPNN calculation as repeated sampling tests, to addressing APRMPS. Finally, with two applied examples of the anti-penetration randomness of metal targets, this paper demonstrates the application procedures for this method, proves the higher efficiency of Dynamic Lifecycle Genetic Algorithm than Genetic Algorithm in optimizing Back Propagation Neural Network and verifies the effectiveness of DLGABPNN-MCS in studying APRMPS. This paper may have some significance in proposing Dynamic Lifecycle Genetic Algorithm—the new, universal heuristic algorithm, and providing the common technical method for the APRMPS study based on DLGABPNN-MCS, in the hope of promoting the application of Dynamic Lifecycle Genetic Algorithm in other optimization problems, and offering reference for the further study of APRMPS or the study of other random problems.</p>

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

Study on the anti-penetration randomness of metal protective structures based on optimized artificial neural network

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年5月13日

查看全文
  • Lan Liu,
  • Weidong Chen,
  • Shengzhuo Lu,
  • Yanchun Yu,
  • Mingwu Sun

摘要

In order to study the Anti-Penetration Randomness of Metal Protective Structures (APRMPS) for the penetration probabilities of Metal Protective Structures under the action of the basic random variables, this paper analyzes the candidates for the basic random variables and the random response of APRMPS, and, on the basis of the improvement of Genetic Algorithm, proposes Dynamic Lifecycle Genetic Algorithm, including its main processes of the optimization of Back Propagation Neural Network. And by adopting the Back Propagation Neural Network optimized by Dynamic Lifecycle Genetic Algorithm (DLGABPNN) as the surrogate model of APRMPS, this paper presents the technical route of DLGABPNN-MCS, the Monte Carlo Simulation with DLGABPNN calculation as repeated sampling tests, to addressing APRMPS. Finally, with two applied examples of the anti-penetration randomness of metal targets, this paper demonstrates the application procedures for this method, proves the higher efficiency of Dynamic Lifecycle Genetic Algorithm than Genetic Algorithm in optimizing Back Propagation Neural Network and verifies the effectiveness of DLGABPNN-MCS in studying APRMPS. This paper may have some significance in proposing Dynamic Lifecycle Genetic Algorithm—the new, universal heuristic algorithm, and providing the common technical method for the APRMPS study based on DLGABPNN-MCS, in the hope of promoting the application of Dynamic Lifecycle Genetic Algorithm in other optimization problems, and offering reference for the further study of APRMPS or the study of other random problems.