Scholarlink

<p>Mechanical press is characterized by strong nonlinearity and parameter fluctuation under the conditions of high-speed stamping and heavy-load operation. Traditional control methods are insufficient in dynamic response and precision maintenance. Focusing on the above problems, this paper constructs a dynamic control model based on fuzzy PID and neural network, and systematically models and verifies the operation process of mechanical press. The dynamic model is established based on the measured operation data, fuzzy rules are introduced to realize PID parameter adjustment, and neural network is used to complete online parameter correction. Comparative experiments were carried out under the conditions of punching pressure of 200–500 kN and punching frequency of 40–60 times /min. The results show that the fusion control is superior to the fixed PID and the single fuzzy PID in tracking accuracy, response time and disturbance recovery, the steady-state error is controlled below 0.15&#xa0;mm, and the adjustment time is kept within 300 ms Relevant conclusions provide engineering reference for intelligent control scheme design of mechanical press.</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

The implementation of fuzzy PID and neural network fusion algorithm in the dynamic control of mechanical presses

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年6月23日

查看全文
  • Yanan Li,
  • Yimin Chu,
  • Ruiyang Yao,
  • Na Sun,
  • Jianzhi Hu

摘要

Mechanical press is characterized by strong nonlinearity and parameter fluctuation under the conditions of high-speed stamping and heavy-load operation. Traditional control methods are insufficient in dynamic response and precision maintenance. Focusing on the above problems, this paper constructs a dynamic control model based on fuzzy PID and neural network, and systematically models and verifies the operation process of mechanical press. The dynamic model is established based on the measured operation data, fuzzy rules are introduced to realize PID parameter adjustment, and neural network is used to complete online parameter correction. Comparative experiments were carried out under the conditions of punching pressure of 200–500 kN and punching frequency of 40–60 times /min. The results show that the fusion control is superior to the fixed PID and the single fuzzy PID in tracking accuracy, response time and disturbance recovery, the steady-state error is controlled below 0.15 mm, and the adjustment time is kept within 300 ms Relevant conclusions provide engineering reference for intelligent control scheme design of mechanical press.