Scholarlink

<p>Cooperative operations of underactuated autonomous underwater vehicles are challenged by limited communication and energy. To address these issues, this paper proposes a dynamic event-triggered distributed optimal formation control method based on an identifier-critic structure. The method has the following features: (1) by reconstructing dynamics and developing a minimum learning parameter-based radial basis function neural network (RBFNN) identifier to solve the completely unknown dynamic problem, and introducing filtered control inputs to avoid the algebraic loop issue; (2) a filtered formation error equation is transformed into an affine-form error dynamic system, thereby achieving distributed overall optimal control. (3) an improved non-quadratic cost function is designed to handle asymmetric input constraints, while a dynamic event-triggering mechanism and a critic RBFNN are used to construct an approximate optimal control law. Additionally, a weight update law without initial weight adjustment is developed. Compared with existing studies, the method optimizes the trade-off between formation control performance and energy consumption, while also reducing computation and communication burdens. Theoretical analysis demonstrates that all signals in the closed-loop system are uniformly ultimately bounded, and the Zeno phenomenon is excluded. Simulation results demonstrate the effectiveness of the method and the advantages in reducing energy consumption, communication, and computational burdens.</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

Dynamic event-triggered distributed neural network constrained optimal formation control of underactuated AUVs under unknown dynamics

  • 期刊论文

  • 发表日期：

    2026年4月8日

查看全文
  • Huibin Gong,
  • Li-Ying Hao,
  • Yuan-Xin Li

摘要

Cooperative operations of underactuated autonomous underwater vehicles are challenged by limited communication and energy. To address these issues, this paper proposes a dynamic event-triggered distributed optimal formation control method based on an identifier-critic structure. The method has the following features: (1) by reconstructing dynamics and developing a minimum learning parameter-based radial basis function neural network (RBFNN) identifier to solve the completely unknown dynamic problem, and introducing filtered control inputs to avoid the algebraic loop issue; (2) a filtered formation error equation is transformed into an affine-form error dynamic system, thereby achieving distributed overall optimal control. (3) an improved non-quadratic cost function is designed to handle asymmetric input constraints, while a dynamic event-triggering mechanism and a critic RBFNN are used to construct an approximate optimal control law. Additionally, a weight update law without initial weight adjustment is developed. Compared with existing studies, the method optimizes the trade-off between formation control performance and energy consumption, while also reducing computation and communication burdens. Theoretical analysis demonstrates that all signals in the closed-loop system are uniformly ultimately bounded, and the Zeno phenomenon is excluded. Simulation results demonstrate the effectiveness of the method and the advantages in reducing energy consumption, communication, and computational burdens.