Scholarlink

<p>Smart manufacturing is a key industry that needs fine control optimization in order to maintain product quality, minimize defects, and maximize energy usage. Nonetheless, the strongly nonlinear and constrained characteristics of most manufacturing processes hinder the applicability of conventional control optimization approaches, i.e., Proportional-Integral-Derivative (PID), which tend to have difficulty dealing with dynamic changes and complexity in systems. A hybrid Nonlinear MPC (NMPC) with Fuzzy PID (NMPC + Fuzzy PID) architecture is introduced for real-time process optimization and automation in smart manufacturing. The method combines NMPC for predictive optimization with Fuzzy PID for addressing nonlinearities and uncertainties. Modeling of nonlinear systems encompasses the dynamic characteristics of manufacturing processes, and fuzzy logic promotes the flexibility of PID control optimization in addressing rapid process dynamics. Experimental results and physical laws were utilized to select and verify the nonlinear models, which were then added to the NMPC structure. Optimization of the control utilizes iterative computation to achieve real-time response while dealing with constraints like pressure, temperature, and material flow rates. The framework was implemented using MATLAB R2023a tested in a production environment, where dimensional accuracy and defect minimization are critical. The results settling time (120&#xa0;s) showed notable enhancements in product quality, with a significant reduction in defects. The system also achieved improvements in energy efficiency compared to traditional control optimization methods. It effectively addresses nonlinearities and constraints, achieving superior performance in smart manufacturing. Future directions include integrating machine learning (ML) for improved model accuracy and exploring scalability for complex, multi-step production processes.</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

Real-time process optimization in smart manufacturing with fuzzy PID and model predictive control

  • 期刊论文

  • 发表日期：

    2025年7月1日

查看全文
  • Hrushikesh Sarangi,
  • Pooja Goel,
  • Dhruv Kumar,
  • A. Anderson,
  • Kuthalingam Venkadeshwaran

摘要

Smart manufacturing is a key industry that needs fine control optimization in order to maintain product quality, minimize defects, and maximize energy usage. Nonetheless, the strongly nonlinear and constrained characteristics of most manufacturing processes hinder the applicability of conventional control optimization approaches, i.e., Proportional-Integral-Derivative (PID), which tend to have difficulty dealing with dynamic changes and complexity in systems. A hybrid Nonlinear MPC (NMPC) with Fuzzy PID (NMPC + Fuzzy PID) architecture is introduced for real-time process optimization and automation in smart manufacturing. The method combines NMPC for predictive optimization with Fuzzy PID for addressing nonlinearities and uncertainties. Modeling of nonlinear systems encompasses the dynamic characteristics of manufacturing processes, and fuzzy logic promotes the flexibility of PID control optimization in addressing rapid process dynamics. Experimental results and physical laws were utilized to select and verify the nonlinear models, which were then added to the NMPC structure. Optimization of the control utilizes iterative computation to achieve real-time response while dealing with constraints like pressure, temperature, and material flow rates. The framework was implemented using MATLAB R2023a tested in a production environment, where dimensional accuracy and defect minimization are critical. The results settling time (120 s) showed notable enhancements in product quality, with a significant reduction in defects. The system also achieved improvements in energy efficiency compared to traditional control optimization methods. It effectively addresses nonlinearities and constraints, achieving superior performance in smart manufacturing. Future directions include integrating machine learning (ML) for improved model accuracy and exploring scalability for complex, multi-step production processes.