Scholarlink

A machine learning-based automation control system has achieved significant results through actual application verification, with a control accuracy of ± 0.15 °C and a system response time of 12 ms, which is 42.3% better than traditional PID control methods. After algorithm optimization and system improvement, the concurrent processing capacity has been increased to 3800 times/second, CPU utilization has been reduced to 55%, and energy efficiency has been improved by 25.7%. The system has been running for 720 h in an industrial setting, processing 360 million data points, with a service availability of 99.999%, demonstrating good engineering application value and promotion prospects.

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

Performance Study of Machine Learning Algorithms in Computer-Controlled Automation Systems

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Weiya Zhang

摘要

A machine learning-based automation control system has achieved significant results through actual application verification, with a control accuracy of ± 0.15 °C and a system response time of 12 ms, which is 42.3% better than traditional PID control methods. After algorithm optimization and system improvement, the concurrent processing capacity has been increased to 3800 times/second, CPU utilization has been reduced to 55%, and energy efficiency has been improved by 25.7%. The system has been running for 720 h in an industrial setting, processing 360 million data points, with a service availability of 99.999%, demonstrating good engineering application value and promotion prospects.