학술논문

Feature Extraction of Fourier Infrared Signals from Pyrolysis Products based on ZCA and PSO
Document Type
Conference
Source
2020 IEEE Congress on Evolutionary Computation (CEC) Evolutionary Computation (CEC), 2020 IEEE Congress on. :1-7 Jul, 2020
Subject
Bioengineering
Computing and Processing
General Topics for Engineers
Power, Energy and Industry Applications
Robotics and Control Systems
Signal Processing and Analysis
Analytical models
Spectral analysis
Noise reduction
Standards
Fourier transforms
Libraries
Correlation
Fourier transform infrared spectroscopy
Pyrolysis products
Particle swarm optimization (PSO)
Wavelet threshold filter
Zero-phase component analysis (ZCA) whitening
Language
Abstract
Fourier transform infrared spectroscopy (FTIR) can provide abundant information for different organic functional groups and chemical bonds, and is therefore widely used in molecular structure analysis, qualitative analysis of mixtures, and quantitative analysis. However, it is difficult to decompose and extract the highly correlated component spectra from complex mixed spectrum, and existing methods for FTIR qualitative analysis has poor recognition effect. For this reason, a spectroscopic analysis model based on zero-phase component analysis (ZCA) whitening and particle swarm optimization (PSO) is proposed in this paper. The model uses wavelet threshold filter to de-noise the mixed spectrum, and uses ZCA whitening to remove the correlation of the component spectra. Finally, PSO is used to solve the component concentration. In order to verify the superiority of the ZCA whitening-based spectral analysis model, simulation studies are carried out on signals made up by twenty kinds of component spectra. The simulation results show that the model not only achieves effective qualitative analysis and quantitative analysis of the main components of the mixed spectrum, but also has better performance than the current mainstream spectrum analysis model.