Scholarlink

<p>As the key evaluation index to protective coatings, hardness and wear rate have to be measured by time and labor consuming test. As an alternative, the predictions by artificial neural network model is an effective way. However, the study about the prediction performance on multi-tasks including coatings hardness and wear rate was rarely reported. Thus, this study introduced dynamic structure (DS) into elastic weight consolidation (EWC) and gradient episodic memory (GEM) so as to establish EWC-DS and GEM-DS models. The prediction performance showed that even after solely retraining data from new task, EWC-DS and GEM-DS models still presented good prediction accuracy on hardness without catastrophic forgetting.</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

Prediction of Coatings Hardness and Wear Rate by a Continual Learning Model with Dynamic Structure

  • 期刊论文

  • 发表日期：

    2025年5月27日

查看全文
  • Qianzhi Wang,
  • Da Lei,
  • Jizhou Kong,
  • Zhifeng Zhou

摘要

As the key evaluation index to protective coatings, hardness and wear rate have to be measured by time and labor consuming test. As an alternative, the predictions by artificial neural network model is an effective way. However, the study about the prediction performance on multi-tasks including coatings hardness and wear rate was rarely reported. Thus, this study introduced dynamic structure (DS) into elastic weight consolidation (EWC) and gradient episodic memory (GEM) so as to establish EWC-DS and GEM-DS models. The prediction performance showed that even after solely retraining data from new task, EWC-DS and GEM-DS models still presented good prediction accuracy on hardness without catastrophic forgetting.