Identification and Sorting of PVC Polymer in Recycling Process by Laser-Induced Breakdown Spectroscopy (LIBS) Combined with Support Vector Machine (SVM) Model

Abstract

In this research, the capability of laser-induced breakdown spectroscopy (LIBS) is evaluated by applying support vector machine (SVM) model to separate polyvinyl chloride (PVC) from other polymers in the recycling process. In our experiment; just single-shot spectrum for each sample was recorded to reach fast LIBS-based technique in a real situation in the recycling factory. Plasma emission of five kinds of waste polymers, including polyethylene (PE), polypropylene (PP), polystyrene (PS), polymethyl methacrylate (PMMA), and polyvinyl chloride (PVC), was recorded in the air atmosphere. Relative intensities of C2/C and N/C were selected as input variables for SVM, and then, Radial Basis and Polynomial kernel function and linear function were applied in SVM to classify recorded data related to different kinds of polymers. With polynomial kernel function of degree 2, polymers were separated correctly with an accuracy of 90.5 %. The results of this research demonstrated that the coupling of LIBS with the non-linear SVM method has great potential to be used for on-line, fast, and accurate classification of polymer samples in the recycling process.