Abstract：The identification of coal categories is crucial for commodity classification,quality inspection,and clean utilization.Near-infrared spectroscopy(NIRS) has gained extensive application and sustained attention in various fields due to its advantages of efficiency,accuracy,environmental friendliness,and strong adaptability.In this study,the near-infrared spectral data of 305 imported coal samples originating from 11 different countries were collected.A feature analysis of the near-infrared spectra was conducted in terms of absorbance magnitude,spectral slope,and distribution of characteristic peaks according to coal categories.The results indicated that the individual indicator could reflect the differences in coal categories,but it was insufficient for the accurate classification.A method that combined near-infrared spectroscopy with continuous wavelet transform-convolutional neural network(CWT-CNN) for the identification of coal categories was proposed.Through the CWT,the near-infrared spectra were transformed from two-dimensional data into three-dimensional feature images,enhancing the spectral resolution and amplifying the extraction of subtle features in spectral curves.The derived three-dimensional feature images were then input into a convolutional neural network discrimination model based on the GoogleNet architecture.The Cross-entropy Loss function was employed as the loss function for the identification of coal categories.During the early stages of model training,the biases and weights of learning rate factor were optimized,and different optimizers were compared and selected.After cross-validation,the average accuracy of training set,validation set and of test set optimized model were 99.69%,96.69% and 96.39%,respectively.
[1] 中华人民共和国国家质量监督检验检疫总局,中国国家标准化管理委员会.GB/T 5751—2009 中国煤炭分类[S].北京:中国标准出版社,2009. [2] 严衍禄,赵龙莲,韩东海,等.近红外光谱分析基础与应用[M].北京:中国轻工业出版社,2005. [3] 洪子云,严承琳,闵红,等.基于近红外光谱判断分析煤种的鉴别研究[J].光谱学与光谱分析,2022,42(9):2800-2806. HONG Ziyun,YAN Chenglin,MIN Hong,et al.Study on identification of coal types based on near-infrared spectroscopy discriminant analysis[J].Spectroscopy and Spectral Analysis,2022,42(9):2800-2806. [4] 宋亮,刘善军,毛亚纯,等.基于可见光-近红外光谱的煤种分类方法[J].东北大学学报(自然科学版),2017,38(10):1473-1476. SONG Liang,LIU Shanjun,MAO Yachun,et al.Classification method of coal types based on visible-near infrared spectroscopy[J].Journal of Northeastern University(Natural Science),2017,38(10):1473-1476. [5] 王雅圣,杨梦,骆志远,等.基于置信学习机与近红外光谱的煤种快速分类方法[J].光谱学与光谱分析,2016,36(6):1685-1689. WANG Yasheng,YANG Meng,LUO Zhiyuan,et al.Rapid classification method for coal types based on confidence learning machine and near-infrared spectroscopy[J].Spectroscopy and Spectral Analysis,2016,36(6):1685-1689. [6] 李彦冬,郝宗波,雷航,等.卷积神经网络研究综述[J].计算机应用,2016,36(9):2508-2515,2565. LI Yandong,HAO Zongbo,LEI Hang,et al.A review of convolutional neural networks[J].Computer Applications,2016,36(9):2508-2515,2565. [7] 廖家慧,杨丰,詹长安,等.结合连续小波变换与生成对抗网络的癫痫发作预测[J].中国生物医学工程学报,2023,42(2):168-179. LIAO Jiahui,YANG Feng,ZHAN Chang′an,et al.Seizure prediction combining continuous wavelet transform with generative adversarial networks[J].Journal of Chinese Biomedical Engineering,2023,42(2):168-179. [8] 吕杭,蒋明峰,李杨,等.基于混合时频域特征的卷积神经网络心律失常分类方法的研究[J].电子学报,2023,51(3):701-711. LÜ Hang,JIANG Mingfeng,LI Yang,et al.Research on convolutional neural network-based arrhythmia classification method using hybrid time-frequency domain features[J].Journal of Electronics,2023,51(3):701-711. [9] 中华人民共和国国家质量监督检验检疫总局,中国国家标准化管理委员会.GB 474—2008煤样的制备方法[S].北京:中国标准出版社,2008. [10] 杨恩,王世博,葛世荣.典型块状煤的可见-近红外光谱特征研究[J].光谱学与光谱分析,2019,39(6):1717-1723. YANG En,WANG Shibo,GE Shirong.Study on the visible-near infrared spectral characteristics of typical lumpy coal[J].Spectroscopy and Spectral Analysis,2019,39(6):1717-1723. [11] Mallat S G.A wavelet tour of signal processing:The sparse way[M].3rd ed.Boston:Elsevier/Academic Press,2009. [12] Lilly J M,Olhede S C.Higher-order properties of analytic wavelets[J].IEEE Transactions on Signal Processing,2009,57(1):146-160. [13] 成智礼,郭汉伟.小波与离散变换理论与工程实践[M].北京:清华大学出版社,2005. [14] Szegedy C,Liu W,Jia Y,et al.Going deeper with convolutions [C]//Proceedings of the 2015 IEEE Conference on Computer Vision and Pattern Recognition.Washington D C:IEEE Computer Society,2015:1-8. [15] Lee C,Gallagher P,Tu Z,et al.Generalizing pooling functions in CNNs:mixed,gated,and tree[J].IEEE Transactions on Pattern Analysis Machine Intelligence,2018,40:863-875. [16] Ioffe S,Szegedy C.Batch normalization: acceleration deep network training by reducing internal covariate shirt [C]//International Conference on International Conference on Machine Learning,2015:448-456. [17] 田青林,郭帮杰,叶发旺,等.一维空洞卷积神经网络的矿物光谱分类[J].光谱学与光谱分析,2022,42(3):873-877. TIAN Qinglin,GUO Bangjie,YE Fawang,et al.Mineral spectral classification using one-dimensional dilated convolutional neural networks[J].Spectroscopy and Spectral Analysis,2022,42(3):873-877. [18] 周非,李阳,范馨月,等.图像分类卷积神经网络的反馈损失计算方法改进[J].小型微型计算机系统,2019,40(7):1532-1537. ZHOU Fei,LI Yang,FAN Xinyue,et al.Improved feedback loss computation method for image classification with convolutional neural networks[J].Journal of Mini and Micro Systems,2019,40(7):1532-1537. [19] Cloutis E A.Quantitative characterization of coal properties using bidirectional diffuse reflectance spectroscopy[J].Fuel,2003(82):2239-2254.
