Gao Wenhan, Han Boyuan, Ye Yanpeng, Cai Yuyao, Feng Jun, Yan Yihui, Liu Yuzhu
State Key Laboratory Cultivation Base of Atmospheric Optoelectronic Detection and Information Fusion, Nanjing University of Information Science & Technology, Nanjing 210044, China; Jiangsu International Joint Laboratory on Meteorological Photonics and Optoelectronic Detection, Jiangsu Collaborative Innovation Center on Atmospheric Environment and Equipment Technology (CICAEET), Nanjing University of Information Science & Technology, Nanjing 210044, China.
Lehrstuhl für Physikalische Chemie, TUM School of Natural Sciences, Technische Universität München, Lichtenbergstraße 4 85748, Garching, Germany.
Waste Manag. 2025 Feb 1;193:135-142. doi: 10.1016/j.wasman.2024.11.044. Epub 2024 Dec 9.
In the modern electronics industry, with the rapid development of technology and the quick turnover of electronic products, the production of electronic waste (e-waste) has also dramatically increased. Among these, discarded capacitors are a significant component of e-waste. These old capacitors not only contain harmful chemicals but are also rich in economically recoverable precious metals like Nb and Ag. This study specifically aims to enhance the classification of discarded capacitors to enable more efficient recycling and resource recovery.Traditional methods of capacitor classification mainly rely on manual identification, which is inefficient and limited in accuracy. To enhance the efficiency and accuracy of classification, this study introduces, for the first time, the combination of Laser-Induced Breakdown Spectroscopy (LIBS) technology and machine learning for the classification of capacitors. The Backpropagation Artificial Neural Network (BP-ANN) algorithms can be trained to automatically identify and classify discarded capacitors. To achieve better performance, we developed a novel algorithm called the Optimized Feature Extraction Variance Algorithm (OFEVA), which addresses the limitations of existing methods by significantly improving the accuracy of the classification model. Compared to training with principal component scores data from traditional Principal Component Analysis (PCA), training with OFEVA achieves higher classification accuracy and computational efficiency.This innovative approach not only helps increase the recycling rate of discarded capacitors and reduce environmental pollution but also provides significant technical support for the reuse of resources, thereby making an important contribution to the fields of environmental protection and resource recycling. In addition, the spectral lines of pure niobium have been calibrated for the first time in this paper, providing important data for further spectroscopic studies.
