文献类型：期刊文献

英文题名：Distributed Reinforcement Learning-Based Optimal Formation Control for UAV-UGV System with Prescribed Performance

作者：Li, Weichen[1]; Fu, Hui[1]; Wu, Qingzhen[1]; Li, Hongpeng[1]; Cui, Yu[1]; Liu, Haitao[2]

机构：[1] School of Mechanical and Electrical Engineering, Guangdong University of Technology, Guangzhou, China; [2] School of Mechanical Engineering, Guangdong Ocean University, Zhanjiang, China

年份：2026

起止页码：1327

外文期刊名：38th Chinese Control and Decision Conference, CCDC 2026

收录：EI(收录号：20262821062998)

语种：英文

外文关键词：Adaptive control systems - Aircraft control - Antennas - Controllers - Errors - Intelligent systems - Learning algorithms - Multi agent systems - Optimal control systems - State estimation - Time varying control systems - Unmanned aerial vehicles (UAV)

外文摘要：This paper introduces the distributed optimal time-varying formation control strategy based on reinforcement learning (RL), which is suitable to heterogeneous multi-agent systems (HMASs) consisting of unmanned aerial vehicles (UAVs) and unmanned ground vehicles (UGVs). For such multiagent systems, the transient performance is considered a key metric. Prescribed performance control can ensure that the control system output error satisfies the desired transient performance. First, the distributed adaptive state observer is proposed to redistribute the novel state messages of the leader to each agent. Second, the transient and steady-state performance of controller is ensured by integrating the prescribed performance function (PPF) into the error transformation. Third, the distributed optimal time-varying formation control strategy based on the actor-critic framework is designed, which can adaptively adjust the controller through reinforcement learning algorithms, and achieve optimal solutions for cost and energy consumption. Simulation results verify the validity and superiority, and the controller error signals are theoretically proven to be bounded within a certain range. ? 2026 IEEE.