Hyperspectral SFIM-RFR Model on Predicting the Total Iron Contents of Iron Ore Powders
GAO Wei1, YANG Ke-ming1*, LI Meng-qian2, LI Yan-ru1, HAN Qian-qian1
1. College of Geoscience and Surveying Engineering, China University of Mining & Technology(Beijing), Beijing 100083, China
2. North China University of Science and Technology, Tangshan 063210, China
Abstract Iron ore is one of the most abundant metallic minerals in the world. Total iron contents is an important index to evaluate the quality of iron ore and iron ore powder, and it has a special significance in iron ore mining, ore dressing, ore smelting and other production links. The traditional chemical methods have the disadvantages of a time-consuming, complex operation, seriously pollution. Therefore, exploring a new method of rapid, effective and pollution-free detection has become a hot spot in mine environment research. Hyperspectral technology has the characteristics of high spectral resolution, continuous curve, no damage, no pollution and accurate detection of characteristics or components of materials. The purpose of this paper is to establisha data evaluation index of spectral feature importance measures (SFIM) and to screen spectral features based on the hyperspectral data of iron ore powder, and then combined with random forest regression (RFR) to establish the SFIM-RFR prediction model and predict the total iron contents of iron ore powder. First, taking Sanyizhuang iron mine in Yangyuan county, Hebei province as a research object, based on the iron concentrate and iron powder tail collected in the research area in November 2018 and March 2019, the first batch of iron ore powder samples in the training group and the testing group and the second batch of iron ore powder samples in the second testing group were made respectively. Spectral data of samples were measured by the ASD Field Spec4 spectrometer. Then, spectral data of the first batch of training group were used in the SFIM-RFR model training, and the total iron contents in the samples of the first batch of the testing group were predicted. Meanwhile, conventional methods, including RFR and linear regression (LR) prediction model, were used to compare and analyze the predicted results of total iron contents in iron ore powder samples. Finally, the spectral data of the second testing group were used to validatethe robustness of the multi-model. The results show that the R-Square values of prediction results of total iron contentsby the SFIM-RFR, RFR and LR models were 0.991 8, 0.988 4, 0.898 7, and RMSE valuesare 0.016 9, 0.020 1, 0.059 6. The results of multi-model prediction overall are good, and the SFIM-RFR model has the minimum error, which indicates the feasibility and effectiveness of this model in predicting the total iron contents of iron ore powder. Moreover, the prediction ability of SFIM-RFR model is better than that of conventional prediction models. The R-Square values of the prediction results of total iron contents by the SFIM-RFR, RFR and LR models are 0.976 8, 0.974 5 and 0.914 0. The RMSE values are 0.034 6, 0.036 2 and 0.071 9, which proves that the prediction ability of the SFIM-RFR model is the best and the robustness of the prediction model is the best.
Key words：Hyperspectral; Total iron contents of iron ore powder; Prediction model; SFIM; Random forest regression
[1] Hu H, Tang Y, Ying H, et al. Talanta, 2014, 125: 425.
[2] MING Hong, REN Li-ping, QIN Ye-qiong, et al(闵 红, 任丽萍, 秦晔琼, 等). Metallurgical Analysis(冶金分析), 2014, 34(4): 21.
[3] TONG Qing-xi, ZHANG Bing, ZHANG Li-fu(童庆禧, 张 兵, 张立福). Journal of Remote Sensing(遥感学报), 2016, 20(5): 689.
[4] DOU Zhi-guo, CUI Li-juan, WU Gao-jie, et al(窦志国, 崔丽娟, 武高洁, 等). Journal of Ecology(生态学杂志), 2018, 37(10): 3163.
[5] ZHANG Zhuo-ran, CHANG Qing-rui, ZHANG Ting-long, et al(张卓然, 常庆瑞, 张廷龙, 等). Journal of Northwest A&F University·Natural Science Edition(西北农林科技大学学报·自然科学版), 2018, 46(11): 1.
[6] SUN Wen-juan, LI Xin-ju(孙问娟，李新举). Journal of Soil and Water Conservation(水土保持学报), 2018, 32(5): 346.
[7] LI Xi-can, ZHAO Geng-xing, CHEN Hang-yan, et al(李西灿, 赵庚星, 陈红艳, 等). Journal of Geomatics Science and Technology(测绘科学技术学报), 2014, 31(6): 593.
[8] SONG Ting-ting, FU Xiu-li, CHEN Yu, et al(宋婷婷, 付秀丽, 陈 玉, 等). Remote Sensing Technology and Application(遥感技术与应用), 2018, 33(1): 88.
[9] Liu Y, Chen Y. International Journal of Remote Sensing, 2012, 33(18): 5954.
[10] Basayigit L, Dedeoglu M, Akgul H. Turkish Jouranal of Agriculture and Forestry, 2015, 39(1): 123.
[11] Guo Y M, Guo L B, Hao Z Q, et al. Journal of Analytical Atomic Spectrometry, 2018, 33(8): 1330.
[12] LI Ying-na, XU Zhi-bin(李颖娜, 徐志彬). Metallurgical Analysis(冶金分析), 2019, 39(1): 35.
[13] HE Qun, WANG Dong, LIU Shan-jun, et al(何 群, 王 东, 刘善军, 等). Metal Mine(金属矿山), 2018, 12: 140.
[14] Breiman L. Machine Learning, 2001, 45(1): 5.
