Chemotaxis-Inspired Control for Multi-Agent Coordination: Formation Control by Two Types of Chemotaxis Controllers
Chemotaxis-Inspired Control for Multi-Agent Coordination: Formation Control by Two Types of Chemotaxis Controllers
复制标题
DOI：
10.1007/s00354-020-00093-0
复制
发表时间：
2020-05
影响因子：
2.6
通讯作者：
Shinsaku Izumi;S. Azuma
中科院分区：
文献类型：
--
作者：
Shinsaku Izumi;S. Azuma
本文研究了受微生物趋化性启发的多智能体系统的控制。趋化性是一种生物现象，其中环境中的生物体被食物吸引但远离毒素。这里解决的问题是编队控制问题，即分布式控制器的设计问题，其中代理的相对位置随着时间的推移变成期望的位置。为了解决这个问题，我们引入了一个性能指标，该指标量化了所需编队的实现程度，并将其分解为可以嵌入分布式控制器中的局部指标。在此基础上，我们提出了受大肠杆菌趋化性启发的编队控制器，其中每个智能体的移动目标都是使用大肠杆菌的趋化性控制器来增加相应的局部性能指数。大肠杆菌。此外，为了提高形成结果的准确性，我们提出了草履虫（P. caudatum）型形成控制器，取代了草履虫的趋化性控制器。上面与 P 的观点相一致。尾部。通过与E的比较证明了有效性。通过数值模拟的大肠杆菌型形成控制器。这一结果意味着各种趋化控制器可以在我们的方法中使用，并且通过选择合适的趋化控制器可以提高所得控制器的性能。
This paper investigates the control of multi-agent systems inspired by chemotaxis of microorganisms. Chemotaxis is a biological phenomenon wherein organisms in an environment are attracted to food but move away from toxins. The problem addressed here is a formation control problem, i.e., a design problem of distributed controllers wherein the relative positions of agents become the desired positions with the progression of time. To solve this problem, we introduce a performance index that quantifies the achieved degree of a desired formation, and decompose it into local indices that can be embedded in the distributed controllers. Based on this, we propose formation controllers inspired by chemotaxis ofEscherichia coli(E. coli), where each agent moves with the aim of increasing the corresponding local performance index using the chemotaxis controller ofE. coli. In addition, to improve the accuracy of the resulting formation, we presentParamecium caudatum(P. caudatum)-type formation controllers by replacing the chemotaxis controller ofE. coliused above with that ofP. caudatum. The effectiveness is demonstrated by a comparison with theE. coli-type formation controllers via numerical simulation. This result implies that various chemotaxis controllers can be used in our method and the performance of the resulting controllers can be improved by choosing an appropriate chemotaxis controller.