Scholarlink

This chapter focuses on an analysis of a study in which a fleet of autonomous surface vehicles collaborate to find and clean plastic debris from aquatic environments. The fleet is divided into two teams with distinct roles. On one side is the scout team, which is in charge of updating and finding the trash items in the environment, thanks to its long-range cameras and high speed. On the other hand, there is the cleaner team, with more basic and short-range cameras to limit economic costs. These are also slower, so they cover half the space in the same time, since they carry the waste collection system. This chapter introduces reinforcement learning and how it can be applied to solve this problem, using a previous work as an example. In addition, an ablation and parameter tuning study of two key components of reinforcement learning is conducted: the reward function and the observation function.

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 Reinforcement Learning and Informative Path Planning: Diving into the Cooperation of Heterogeneous Aquatic Surface Vehicles

  • 章节

  • 发表日期：

    2026年

查看全文
  • Alejandro Mendoza Barrionuevo,
  • Samuel Yanes Luis,
  • Daniel Gutiérrez Reina,
  • Sergio L. Toral Marín

摘要

This chapter focuses on an analysis of a study in which a fleet of autonomous surface vehicles collaborate to find and clean plastic debris from aquatic environments. The fleet is divided into two teams with distinct roles. On one side is the scout team, which is in charge of updating and finding the trash items in the environment, thanks to its long-range cameras and high speed. On the other hand, there is the cleaner team, with more basic and short-range cameras to limit economic costs. These are also slower, so they cover half the space in the same time, since they carry the waste collection system. This chapter introduces reinforcement learning and how it can be applied to solve this problem, using a previous work as an example. In addition, an ablation and parameter tuning study of two key components of reinforcement learning is conducted: the reward function and the observation function.