Experimental Study On Classification And Sorting Mode Of Mixed Plastics Based On Near Infrared Technology

Plastic products have penetrated into every aspect of our lives,and waste plastic products have caused serious environmental problems.In the field of classification and identification of organic compounds,the range of analysis of near-infrared spectroscopy can cover almost all organic compounds.Based on the near-infrared spectroscopy technology,this subject has carried out related research on the analysis of the types of waste household plastics in order to achieve accurate results.Quickly identify the purpose of the types of plastics for used household appliances.The PCA algorithm designed in this paper optimizes the data processing structure,improves the chemometrics feature extraction algorithm,selects the representative principal component dimension,and expands the nonlinear relationship between the data,making each principal component more representative and effective.The effect of interference factors such as noise generated during spectral acquisition,band overlap and baseline drift on the experimental results.In the experimental verification stage,plastic tablets of different batches of used household electrical appliances were collected as experimental samples,and the PCA algorithm was used to analyze the data and obtain dimensionality reduction data.The clustering analysis method was used to classify and identify the sample data after dimension reduction,and the PCA algorithm was verified experimentally based on the obtained experimental results.According to the comparison results,the algorithm structure is optimized to improve the recognition accuracy.The experimental verification of ABS,PP,PAC/PPT,PC/ABS,PS,PVC and nylon samples collected in Haier New Materials Technology Co.,Ltd.finally proved that the dimensionality reduction data obtained by the PCA algorithm designed by this subject was clustered.After the analysis,the recognition accuracy can reach 91%,which basically meets the experimental needs of the subject.Theinfrared spectra of two domestic and foreign spectroscopic instruments are analyzed and compared,and the spectral equipment with higher cost performance and more suitable for the subject is selected.In addition,the theoretical analysis and experimental verification of the electromagnetic vibration feeding device are carried out to study the performance parameters and design standards of the feeding device,so that the material is more easily separated from the spectral collection and the blowing.Through theoretical analysis,algorithm optimization and experimental comprehensive analysis,this paper optimizes the algorithm of identification process,provides reference for plastic recycling technology of used household electrical appliances,reduces the cost and cycle of sorting,and effectively improves the sorting ability of sorting equipment.Sorting precision,promote the promotion and application of plastic near-infrared sorting equipment.