Scholarlink

<p>In the actual production process of gears often because of the selection of heat treatment parameters is unreasonable and can not accurately achieve the small deformation, high precision, less grinding machining allowance heat treatment sample requirements, there are uneven distribution of carburized layer, surface hardness, hardness of the heart can not meet the requirements of the indicators. At the present stage, the method of multi-parameter multi-level combination test block trial production is often used, but its production cycle is long, and the waste of human and material resources is serious. In this study, with the help of machine learning, a support vector machine prediction model of gear tissue distribution is constructed based on heat treatment parameters, and the radial basis functions kernel function is selected as the kernel function of the support vector machine to improve the accuracy of model prediction by optimizing the kernel parameters. The root mean square error value of the final model is 3.16%, and the coefficient of determination is 0.993. The results show that the method of this paper can accurately and efficiently predict the heat treatment results of gears, and save the manufacturing cycle and cost. The precise control of hardness, carburization layer distribution pattern and metallographic organization of ultra-high-strength steel gears can be realized in actual production.</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

Heat treatment control technology of high-strength steel gears based on support vector machine

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年3月5日

查看全文
  • Yanzhong Wang,
  • Libin Zhang,
  • Yulu Su,
  • Hai Liu,
  • HaiLong Yang,
  • Yanyan Chen

摘要

In the actual production process of gears often because of the selection of heat treatment parameters is unreasonable and can not accurately achieve the small deformation, high precision, less grinding machining allowance heat treatment sample requirements, there are uneven distribution of carburized layer, surface hardness, hardness of the heart can not meet the requirements of the indicators. At the present stage, the method of multi-parameter multi-level combination test block trial production is often used, but its production cycle is long, and the waste of human and material resources is serious. In this study, with the help of machine learning, a support vector machine prediction model of gear tissue distribution is constructed based on heat treatment parameters, and the radial basis functions kernel function is selected as the kernel function of the support vector machine to improve the accuracy of model prediction by optimizing the kernel parameters. The root mean square error value of the final model is 3.16%, and the coefficient of determination is 0.993. The results show that the method of this paper can accurately and efficiently predict the heat treatment results of gears, and save the manufacturing cycle and cost. The precise control of hardness, carburization layer distribution pattern and metallographic organization of ultra-high-strength steel gears can be realized in actual production.