论文部分内容阅读
红外光谱是进行化合物定性及定量分析的重要依据,通常需要大量的数据才能较准确地反映所测化合物的结构特征.对应于众多的化合物,红外光谱的数据量非常庞大.因而,在保证红外光谱主要特征基本不变的前提下,如何对红外光谱进行压缩,较大地减少数据量,进而改善红外光谱的存储、检索及处理等方式是一项很有意义的研究工作.小波神经网络(Wavelet neural network),简称小波网络,是基于小波分析所构造的一种新的神经网络模型,目前在化学界尚未见到介绍和应用.本文将其应用于聚苯乙烯薄膜红外光谱的压缩表达.结果表明,小波网络在大量压缩数据的同时,能够很好地恢复原有红外光谱,特别是能够较准确地反映吸收峰的位置和强度.
Infrared spectroscopy is an important basis for qualitative and quantitative analysis of compounds, and often requires a large amount of data to accurately reflect the structural characteristics of the tested compounds.According to many compounds, the amount of data of infrared spectroscopy is very large.Therefore, The main characteristics of the same basic premise, how to compress the infrared spectrum, a large reduction in the amount of data, and thus improve the infrared spectrum storage, retrieval and processing methods is a very significant work.Wavelet neural network (Wavelet neural The wavelet network is a new neural network model based on wavelet analysis and has not been introduced and applied in the field of chemistry at present.This paper applies it to the compressive expression of polystyrene film infrared spectrum.The results show that The wavelet network can restore the original infrared spectrum while compressing a large amount of data, and in particular, can better reflect the position and intensity of the absorption peak.