Scholarlink

In this study, an evaporation dataset of dodecane fuel was experimentally established and modeled for prediction using neural networks. The data was obtained by doing 27 evaporation rates at 3 volumes and 9 temperatures first. From the evaporation principle analysis 8 parameters were obtained then verified their two by two correlation using Pearson correlation to prove their good input parameters. Finally, BP and SVM algorithms were modeled to improve the accuracy by changing the model parameters, and the error comparison method was used to reflect more intuitively the goodness of the model prediction.

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

Neural Network-Based Prediction of the Average Evaporation Rate of a Single Droplet of Dodecane

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Lei Liu,
  • Genzhu Jiang,
  • X. R. Wang,
  • Juxiang Tao,
  • Rui Chen,
  • Yu Wang,
  • H. D. Dai

摘要

In this study, an evaporation dataset of dodecane fuel was experimentally established and modeled for prediction using neural networks. The data was obtained by doing 27 evaporation rates at 3 volumes and 9 temperatures first. From the evaporation principle analysis 8 parameters were obtained then verified their two by two correlation using Pearson correlation to prove their good input parameters. Finally, BP and SVM algorithms were modeled to improve the accuracy by changing the model parameters, and the error comparison method was used to reflect more intuitively the goodness of the model prediction.