In this article, we focus on the cooperative state estimation problem of a multi-agent system. Each agent is equipped with absolute and relative measurements. The purpose of this research is to make each agent generate its own state estimation with only local measurement information and local communication with neighborhood agents using Set Membership Filter(SMF). To handle this problem, we analyzed centralized SMF framework as a benchmark of distributed SMF and propose a finite-horizon method called OIT-Inspired centralized constrained zonotopic algorithm. Moreover, we put forward a distributed Set Membership Filtering(SMFing) framework and develop a distributed constained zonotopic algorithm. Finally, simulation verified our theoretical results, that our proposed algorithms can effectively estimate the state of each agent.

翻译：本文聚焦于多智能体系统的协同状态估计问题。每个智能体配备绝对与相对测量信息。研究目标在于使每个智能体仅利用局部测量信息及与邻域智能体的局部通信，借助集员滤波器生成自身状态估计。为解决该问题，我们以集中式SMF框架作为分布式SMF的基准进行分析，并提出一种名为OIT-Inspired有限时域集中式约束zonotopic算法。此外，我们构建了分布式集员滤波框架，并发展出相应的分布式约束zonotopic算法。最后，仿真验证了理论结果，表明所提算法能有效估计每个智能体的状态。