Scholarlink

<p>The effective generation of decision control rules for intelligent decision-making and behavior planning in Autonomous Underwater Vehicles (AUVs) is crucial. However, traditional decision control rules often rely on manual design, leading to complications and delays. To address these challenges, this chapter proposes a Logic-Optimization Decision-Making Behavior Tree (Logic-DBT) algorithm specifically for multi-AUV. The algorithm utilizes AUV simulations to generate behavior planning data, incorporating a decision tree learning mechanism. Continuous attributes are discretized to facilitate the design of a decision tree for cooperative behavior planning among AUVs. A ternary logic structure is established, with control elements at its core, integrating state judgment and action elements. This structure is then used to create a Decision-Making Behavior Tree (DBT) based on Decision Trees (DTs) by combining temporal logic and constraint modeling. The decision behavior tree’s structure is simplified through a logical operation optimization method, ensuring operational simplicity, timeliness, and reachability. Simulation experiments on automatic behavior control for multi-AUV cooperative tasks demonstrate that the Logic-DBT algorithm provides significant advantages in terms of rapid behavior control response, precise timing logic, and safe, effective operations.</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

Logic-optimization behavior tree algorithm for enhanced autonomous underwater vehicle cooperative decision-making

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年8月1日

查看全文
  • Dan Yu,
  • Changjian Lin,
  • Nan Liu

摘要

The effective generation of decision control rules for intelligent decision-making and behavior planning in Autonomous Underwater Vehicles (AUVs) is crucial. However, traditional decision control rules often rely on manual design, leading to complications and delays. To address these challenges, this chapter proposes a Logic-Optimization Decision-Making Behavior Tree (Logic-DBT) algorithm specifically for multi-AUV. The algorithm utilizes AUV simulations to generate behavior planning data, incorporating a decision tree learning mechanism. Continuous attributes are discretized to facilitate the design of a decision tree for cooperative behavior planning among AUVs. A ternary logic structure is established, with control elements at its core, integrating state judgment and action elements. This structure is then used to create a Decision-Making Behavior Tree (DBT) based on Decision Trees (DTs) by combining temporal logic and constraint modeling. The decision behavior tree’s structure is simplified through a logical operation optimization method, ensuring operational simplicity, timeliness, and reachability. Simulation experiments on automatic behavior control for multi-AUV cooperative tasks demonstrate that the Logic-DBT algorithm provides significant advantages in terms of rapid behavior control response, precise timing logic, and safe, effective operations.