Scholarlink

This paper presents a data-driven approach for identification and control of a DC motor using neural networks in a nonlinear model predictive control framework. First, input-output data are measured in diverse operating range, and used to identify a neural state-space model that gives a good approximation of the dynamics of the motor. The learned model is then embedded in an MPC scheme, enabling the implementation of effective speed regulation. These results prove the validity of using neural identification within control design, while opening paths for next-generation intelligent control schemes. In summary, this is a flexible, high-performance approach which can be readily applied to several industrial and robotic applications.

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

Neural Network–Based DC Motor Identification and Nonlinear MPC Control

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Mokhlis Salah-Eddine,
  • El Haouti Kawtar,
  • Elbaraka Saad,
  • El Assad Nouhaila,
  • Rhouzali Mohamed,
  • Barra Adil,
  • Bensassi Bahloul

摘要

This paper presents a data-driven approach for identification and control of a DC motor using neural networks in a nonlinear model predictive control framework. First, input-output data are measured in diverse operating range, and used to identify a neural state-space model that gives a good approximation of the dynamics of the motor. The learned model is then embedded in an MPC scheme, enabling the implementation of effective speed regulation. These results prove the validity of using neural identification within control design, while opening paths for next-generation intelligent control schemes. In summary, this is a flexible, high-performance approach which can be readily applied to several industrial and robotic applications.