.png){kind=icon}
.png){kind=icon}
Return
Research on rapid water tracing by substrate interface optimization coupling LIBS-machine learning
.png){kind=icon}
PRE
AI
DOI：10.1016/j.microc.2025.116492.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
• Rapid sewage tracing (LIBS+ML) with 3 substrates (PMMA, superhydrophobic cone, Ti foam) fixtraditional LIBS spectral instability. • High-precision 5-region water sample identification: PCA + PSO-LSSVM/SVM/BPNN; PMMA-based PSO-PCA-LSSVM: 99.0 % recognition rate. • PMMA: 84.8 % cumul. contrib. (400 PCs) > SH cone/Ti foam (62.1 %/59.7 %), low 200–500 nm spectral interference. • Supports Baiyangdian sewage origin identification, advancing LIBS in rapid urban sewage tracing and trace element quantitative analysis.