Scholarlink

One technique within additive manufacturing (AM) is wire arc additive manufacturing (WAAM), which uses an electric arc to melt wire material. It builds and constructs a portion by depositing material layer by layer. The current work tried to investigate the accuracy of surface roughness (Ra) for a stainless-steel (SS316L) structure, fabricated by WAAM using a classification-based machine learning model (MLM). The three input parameters and one output parameter were considered to generate a matrix for training and testing. However, a total of 27 experimental data sets were chosen, and the two types of MLM, such as K-nearest neighbours (KNN) and Random Forest (RF), were evaluated. Both MLM accuracy was assessed by mean square error (MSE) and percentage of regression coefficient (R2) for training and testing data sets. It is noticed that both MLMs R2 values are more than 91%, yet the KNN model exhibited the highest accuracy of 99.35%, and the MSE is 0.014.

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

Enhance the Machine Learning Model (MLM) Accuracy of Stainless-Steel Structure Deposition by Wire Arc Additive Manufacturing (WAAM) Process

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Shatarupa Biswas,
  • Manish Soni,
  • Vishal Kumar,
  • Amitava Mandal

摘要

One technique within additive manufacturing (AM) is wire arc additive manufacturing (WAAM), which uses an electric arc to melt wire material. It builds and constructs a portion by depositing material layer by layer. The current work tried to investigate the accuracy of surface roughness (Ra) for a stainless-steel (SS316L) structure, fabricated by WAAM using a classification-based machine learning model (MLM). The three input parameters and one output parameter were considered to generate a matrix for training and testing. However, a total of 27 experimental data sets were chosen, and the two types of MLM, such as K-nearest neighbours (KNN) and Random Forest (RF), were evaluated. Both MLM accuracy was assessed by mean square error (MSE) and percentage of regression coefficient (R2) for training and testing data sets. It is noticed that both MLMs R2 values are more than 91%, yet the KNN model exhibited the highest accuracy of 99.35%, and the MSE is 0.014.