Scholarlink

<p>This paper presents a novel methodology that extends the concepts of Gibbs Random Fields (GRFs) to swarm robotics, facilitating decentralized control of diverse swarm behaviors using only local information. In this approach, a GRF is a probabilistic graph model that describes robot interactions through a Gibbs distribution, guiding each robot to converge to a desired global behavior via locally sampled velocity commands using Markov Chain Monte Carlo (MCMC). This decentralized framework demonstrates scalability, robustness, and flexibility as robots respond autonomously to environmental changes and individual failures. To demonstrate the methodology’s versatility, we address three primary challenges in swarm robotics: flocking-segregative, cooperative transport, and pattern formation behaviors. Simulations and real-world experiments show the method’s adaptability, with the flocking-segregative strategy achieving cohesive navigation and segregation, cooperative transport handling varied object shapes and sizes, and pattern formation creating structured configurations, including chain-like assemblies suitable for modular robotic applications. Overall, our approach highlights the potential of GRFs in swarm robotics, providing a versatile foundation for developing intricate, scalable, and autonomous swarm behaviors.</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

Decentralized Emergent Behavior Design in Robotic Swarms Using Gibbs Random Fields

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年3月25日

查看全文
  • Paulo Rezeck,
  • Luiz Chaimowicz

摘要

This paper presents a novel methodology that extends the concepts of Gibbs Random Fields (GRFs) to swarm robotics, facilitating decentralized control of diverse swarm behaviors using only local information. In this approach, a GRF is a probabilistic graph model that describes robot interactions through a Gibbs distribution, guiding each robot to converge to a desired global behavior via locally sampled velocity commands using Markov Chain Monte Carlo (MCMC). This decentralized framework demonstrates scalability, robustness, and flexibility as robots respond autonomously to environmental changes and individual failures. To demonstrate the methodology’s versatility, we address three primary challenges in swarm robotics: flocking-segregative, cooperative transport, and pattern formation behaviors. Simulations and real-world experiments show the method’s adaptability, with the flocking-segregative strategy achieving cohesive navigation and segregation, cooperative transport handling varied object shapes and sizes, and pattern formation creating structured configurations, including chain-like assemblies suitable for modular robotic applications. Overall, our approach highlights the potential of GRFs in swarm robotics, providing a versatile foundation for developing intricate, scalable, and autonomous swarm behaviors.