Scholarlink

<p>This article proposes a data-driven sliding mode control strategy to address the prescribed performance control problem in heterogeneous multi-agent systems. The scheme transforms the constrained distributed measurement error into an equivalent unconstrained form. Subsequently, the sliding mode controller with prescribed performance is constructed leveraging the compact-form dynamic linearization (CFDL) technique. Besides, the distributed measurement error is consistently constrained within predetermined asymmetric boundaries. Additionally, a boundary condition for the prescribed steady-state performance is provided, revealing that the predefined convergence rate and the topological structure are the primary factors influencing the minimum steady-state value. In contrast to the conventional method, simulation results indicate that the presented approach reduces the mean square error by factors of 2.29 and 3.09 for constant and time-varying signals respectively. Finally, the practicality of the scheme is validated leveraging the multi DC motor system, demonstrating that the mean square error can be consistently maintained at around 10<sup>−3</sup>.</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

Distributed Data-driven Sliding Mode Control for Nonlinear Multi-agent Systems Under Prescribed Performance

  • 期刊论文

  • 发表日期：

    2025年7月14日

查看全文
  • Yijie Yang,
  • Dong Liu,
  • Xin Wang,
  • Zhujun Wang

摘要

This article proposes a data-driven sliding mode control strategy to address the prescribed performance control problem in heterogeneous multi-agent systems. The scheme transforms the constrained distributed measurement error into an equivalent unconstrained form. Subsequently, the sliding mode controller with prescribed performance is constructed leveraging the compact-form dynamic linearization (CFDL) technique. Besides, the distributed measurement error is consistently constrained within predetermined asymmetric boundaries. Additionally, a boundary condition for the prescribed steady-state performance is provided, revealing that the predefined convergence rate and the topological structure are the primary factors influencing the minimum steady-state value. In contrast to the conventional method, simulation results indicate that the presented approach reduces the mean square error by factors of 2.29 and 3.09 for constant and time-varying signals respectively. Finally, the practicality of the scheme is validated leveraging the multi DC motor system, demonstrating that the mean square error can be consistently maintained at around 10−3.