Related Experiment Video
Updated: Jul 26, 2025
Three-Dimensionally Printed Microfluidic Cross-flow System for Ultrafiltration/Nanofiltration Membrane Performance Testing
Published on: February 13, 2016
Electrospun nanofiber membrane diameter prediction using a combined response surface methodology and machine learning
Md Nahid Pervez1,2, Wan Sieng Yeo3, Mst Monira Rahman Mishu4
1Hubei Provincial Engineering Laboratory for Clean Production and High Value Utilization of Bio-Based Textile Materials, Wuhan Textile University, Wuhan, 430200, China.
This study introduces a machine learning approach to optimize electrospinning, accurately predicting nanofiber membrane diameter. The developed Locally Weighted Kernel Partial Least Squares Regression model offers a sustainable and effective solution for electrospinning processes.
Area of Science:
- Materials Science and Engineering
- Chemical Engineering
- Computational Modeling
Background:
- Electrospinning is a widely studied technique for nanofiber production, but simulation studies remain scarce.
- Optimizing electrospinning parameters is crucial for controlling nanofiber characteristics, such as diameter.
- Existing simulation methods often lack the predictive accuracy needed for efficient process design.
Purpose of the Study:
- To develop a sustainable and effective electrospinning process using a combination of experimental design and machine learning.
- To accurately predict the diameter of electrospun nanofiber membranes.
- To compare the performance of a novel Locally Weighted Kernel Partial Least Squares Regression (LW-KPLSR) model against other regression techniques.
Main Methods:
- Developed a Locally Weighted Kernel Partial Least Squares Regression (LW-KPLSR) model integrated with Response Surface Methodology (RSM).
- Evaluated model accuracy using Root Mean Square Error (RMSE), Mean Absolute Error (MAE), and Coefficient of Determination (R²).
- Compared LW-KPLSR with Principal Component Regression (PCR), Locally Weighted Partial Least Squares Regression (LW-PLSR), Partial Least Square Regression (PLSR), Fuzzy Modelling, and Least Square Support Vector Regression (LSSVR).
Main Results:
- The LW-KPLSR model demonstrated superior performance in predicting electrospun nanofiber membrane diameter compared to all other tested models.
- LW-KPLSR achieved significantly lower RMSE and MAE values, indicating higher prediction accuracy.
- The model attained a high coefficient of determination (R²) of 0.9989, signifying excellent model fit.
Conclusions:
- The developed LW-KPLSR model provides a highly accurate and reliable method for predicting nanofiber diameter in electrospinning.
- This machine learning approach enhances the sustainability and effectiveness of the electrospinning process.
- The findings offer a valuable tool for optimizing electrospun material production and design.
