Research Advances Towards Machine Learning‐Driven Fabrication of Electrospun Films With Excellent Performances

ABSTRACT

Electrospinning technology has been effectively employed to produce nanofiber films with superior performance. While the microstructure and performance of the films are highly dependent on the process parameters with complex nonlinear interactions. With the development of artificial intelligence, machine learning (ML) method has been widely used to establish the relationship among process parameters, structure and performance of electrospun films. This review analyzes the process parameters those would influence the electrospinning process and the workflows and evaluation approaches of ML method in the field of electrospinning in detail. Then an in‐depth discussion of the current applications of ML in the field of electrospinning, including performance prediction and optimization, as well as intelligent detection and classification of defects in nanofibers is provided. The review also analyzes the closed‐loop control of electrospinning in which defect information identified by ML is used to provide real‐time feedback for adjusting process parameters, thereby achieving automatic elimination of defects. Furthermore, the specific application cases of ML‐based electrospinning inverse design are summarized to elaborate the approaches of precisely determining the process parameters based on the microstructure and performance requirements of the materials. Finally, the paper summarizes the persistent challenges and offers a forward‐looking perspective on the developmental directions in ML‐driven electrospinning, with the goal of providing a reference for future research and promoting the practical application of the advanced electrospinning technologies.