Scholarlink

<p>To address the issue of poor synchronization accuracy in multi-motor synchronization control systems subject to load disturbances, parameter variations, and coupling effects, this article proposes a composite control method that combines neural network minimum parameter learning adaptive sliding mode control (NN-MPLM-ASMC) and dynamic gain deviation coupling control (DGDCC). First, weight upper bound estimation is employed to replace the node-by-node weight adjustment of traditional radial basis function neural networks (RBFNN). The NN-MPLM-ASMC algorithm reduces the computational burden, enhances the tracking accuracy and dynamic response speed of single-motor systems, and effectively suppresses chattering. Second, a model reference identification method (MRIM) is introduced to identify the moment of inertia of the permanent magnet synchronous motor (PMSM) in real time. This mechanism enables dynamic adjustment of the velocity compensator’s coupling gain, thereby resolving synchronization error accumulation caused by fixed gains in deviation coupling control (DCC). Finally, the experimental results establish that the introduced strategy elevates both the tracking performance and dynamic responsiveness of the system, while drastically improving multi-motor synchronization accuracy, which suggests strong prospects for practical utilization.</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

Multi-motor synchronization control based on neural network adaptive sliding mode and dynamic gain

  • 期刊论文

  • 发表日期：

    2026年6月3日

查看全文
  • Lei Wang,
  • Ximei Zhao,
  • Hongyan Jin

摘要

To address the issue of poor synchronization accuracy in multi-motor synchronization control systems subject to load disturbances, parameter variations, and coupling effects, this article proposes a composite control method that combines neural network minimum parameter learning adaptive sliding mode control (NN-MPLM-ASMC) and dynamic gain deviation coupling control (DGDCC). First, weight upper bound estimation is employed to replace the node-by-node weight adjustment of traditional radial basis function neural networks (RBFNN). The NN-MPLM-ASMC algorithm reduces the computational burden, enhances the tracking accuracy and dynamic response speed of single-motor systems, and effectively suppresses chattering. Second, a model reference identification method (MRIM) is introduced to identify the moment of inertia of the permanent magnet synchronous motor (PMSM) in real time. This mechanism enables dynamic adjustment of the velocity compensator’s coupling gain, thereby resolving synchronization error accumulation caused by fixed gains in deviation coupling control (DCC). Finally, the experimental results establish that the introduced strategy elevates both the tracking performance and dynamic responsiveness of the system, while drastically improving multi-motor synchronization accuracy, which suggests strong prospects for practical utilization.