Scholarlink

Abstract <p>The feasibility of using machine learning algorithms to pre-smooth the current-voltage characteristic (IVC) of a Langmuir probe before numerical differentiation to obtain the electron energy distribution function (EEDF) was investigated. It was found that the EEDF shape depends strongly on the IVC region used to train the model. Excluding the IVC region corresponding to the electron-saturation branch from the training process reproduces a local EEDF maximum at ~17 eV. Good agreement with the EEDF obtained by the independent splicing method in the energy range of 0 to 23 eV was demonstrated.</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

Application of Machine Learning Algorithms to Obtain the EEDF from the I–V Characteristics of a Langmuir Probe

  • 期刊论文

  • 发表日期：

    2026年4月1日

查看全文
  • A. V. Bernatskiy,
  • I. I. Draganov,
  • P. G. Ivanov,
  • V. V. Lagunov,
  • V. N. Ochkin

摘要

Abstract

The feasibility of using machine learning algorithms to pre-smooth the current-voltage characteristic (IVC) of a Langmuir probe before numerical differentiation to obtain the electron energy distribution function (EEDF) was investigated. It was found that the EEDF shape depends strongly on the IVC region used to train the model. Excluding the IVC region corresponding to the electron-saturation branch from the training process reproduces a local EEDF maximum at ~17 eV. Good agreement with the EEDF obtained by the independent splicing method in the energy range of 0 to 23 eV was demonstrated.