基于XGBoost的电磁阀滤网缺陷检测系统设计

打开文本图片集

关键词： 电磁阀滤网； 缺陷检测； XGBoost 算法； 随机森林； 特征优选； 超参数寻优中图分类号：TN911.73⁃34；TP391 文献标识码： A 文章编号：1004⁃373X（2025）12⁃0013⁃06

Abstract：In order to make up for the gap of the automated detection of surface defects in automotive electromagnetic valve filters and to solve the problems such as low detection accuracy and missed defects， an XGBoost⁃based surface defect detection system for electromagnetic valve filter screen is designed to effectively identify the holes and loose wire defects in the electromagnetic valve filter screen. In the system， the filter images are filtered and enhanced， and then the filter area and its plastic frame area are located to extract the fabric area of the filter. The grayscale and texture features of the fabric area are extracted， and the feature reduction algorithm based on random forests is used to conduct the feature optimization. An XGBoost model is constructed to recognize and classify defects. A two⁃fold and multi⁃stage hyperparameter optimization strategy is used for parameter optimization to improve the model effect. The experimental results show that the designed system can accurately identify and classify surface defects in the filters， achieving higher accuracy compared to traditional machine learning detection methods.

Keywords：electromagnetic valve filter screen; defect detection; XGBoost algorithm; random forest; feature optimization; hyper⁃parameter optimization

0 引 言

电磁阀是智能汽车线控制动系统的核心部件，其中电磁阀滤网负责拦截制动液中的杂质，保护阀芯和密封件。（剩余7202字）