摘要
为了实现塑料的分类回收,需要对塑料进行快速准确的鉴别。收集了丙烯腈-丁二烯-苯乙烯(ABS)、聚丙烯(PP)、聚乙烯(PE)、聚对苯二甲酸乙二醇酯(PET)、聚苯乙烯(PS)、聚氯乙烯(PVC)、聚碳酸酯(PC)等7种常用的塑料,利用近红外光谱仪分别测得其反射光谱,应用主成分分析和反向传播(BP)神经网络建立模型进行鉴别。首先利用主成分分析提取光谱的特征信息,前8个主成分的累计贡献率达到94.367%,包含了原始光谱的主要信息,将这8个主成分作为BP神经网络的输入,7种塑料的种类作为输出,建立三层BP神经网络模型。每种塑料各30个样本共210个用来训练神经网络模型,各10个共70个用来预测,预测结果准确率达98.571%,能够有效鉴别常用塑料。
In order to achieve the classification in plastic recycling, it is needed to identify plastics quickly and accurately. In this paper, we collected seven kinds of common plastics such as acrylonitrile- butadiene-styrene copolymer （ABS）, polypropylene （PP）, polyethylene （PE）, polyethylene terephthalate （PET） polystyrene （ PS）, polyvinyl chloride （PVC） and polycarbonate （ PC）, and used near-infrared spectroscopy to obtain their reflectance spectrum establish the identification model by principal component analysis （PCA） and back-propagation BP neural network. PCA was firstly applied to extract the characteristics of spectrum, and the first eight principal components （PCs） contained the main information of the original spectrum with cumulative contribution of 94. 367%. These eight PCs were taken as the input of BP neural network along with the category of seven kinds of plastics as the output to build a three-layer BP neural network model. Thirty samples per each kind of plastic-210 samples in total-were used to train the neural network model, and 70 samples in total with 10 samples per each kind were selected as prediction set, the result of which leads to 98. 571% of accuracy rate, indicating that it can be effective for the identification of common plastics.
出处
《塑料工业》
CAS
CSCD
北大核心
2016年第12期124-127,137,共5页
China Plastics Industry
关键词
近红外光谱
塑料
主成分分析
反向传播神经网络
鉴别
Near-infrared Spectrum
Plastics
Principal Component Analysis
Back-propagationNeural Network
Identification
