Inversion of Cadmium Content in Agriculture Soil Based on SGA-RF Algorithm

被引：0

作者：

Wang X. ^{[1
,2
]}

Chen J. ^{[2
]}

Zheng X. ^{[1
]}

Zhu C. ^{[3
]}

Wang X. ^{[1
,2
]}

Shan C. ^{[5
]}

机构：

[1] Key Laboratory of Marine Environmental Science and Ecology, Ministry of Education, Ocean University of China, Qingdao

[2] Science and Information College, Qingdao Agricultural University, Qingdao

[3] Project Management Department, China Unicom Ji'nan Software Research Institute, Ji'nan

[4] Information Engineering and Automation Department, Shanxi Institute of Technology, Yangquan

[5] The Environmental Monitoring Center of North China Sea, Qingdao

来源：

Zheng, Xilai (zhxilai@ouc.edu.cn) | 2018年 / Chinese Society of Agricultural Machinery卷 / 49期

关键词：

Agriculture soil; Cadmium content; Characteristic wavelength selection; Genetic algorithm; Random forest; Spearman's rank correlation analysis;

D O I：

10.6041/j.issn.1000-1298.2018.10.029

中图分类号：

学科分类号：

摘要：

In the field of hyperspectral detection on heavy metal pollution levels in agricultural soils, the accuracy and stability of hyperspectral inversion model for soil cadmium were seriously affected by the high dimensional and high redundancy characteristics in visible/NIR spectra. In order to solve the above problems, Spearman's rank correlation analysis-based genetic algorithm by using random forest (SGA-RF) was proposed to select the characteristic wavelength from hyperspectral data. On the first-layer of feature selection stage, Spearman correlation analysis-based feature selection method was applied to remove redundancy between all spectra features and retain the characteristic wavelength which was the most relevant to the cadmium content. On the second-layer of feature selection stage, a new fitness function based on random forest was proposed, which perfectly combined the strong global search ability of genetic algorithm and the high inversion ability of random forest. With the proposed fitness function to evaluate the viability of individuals, the distinguishing ability between similar individuals was improved and a subset of optimal spectra feature set with minimum redundancy and maximum differentiation were obtained. In order to verify the validity of the proposed algorithm, totally 124 representative soil samples collected from the Dagu River Basin were chosen as samples. The optimal feature subset which contained 37 sensitive wavelengths was chosen and used to build soil available cadmium content inversion model, and its performance was compared with that of current feature selection methods. Results indicated that the minimum numbers of wavelength features was selected and meanwhile the prediction performance had lower predictive root mean square error of 0.060 1, higher correlation coefficient of 0.950 2 and residual predictive deviation of 2.03. As an important step for the quantitative inversion of cadmium concentration by using visible/NIR spectra, the research could provide some theoretical basis for monitoring soil heavy metal pollution. © 2018, Chinese Society of Agricultural Machinery. All right reserved.

引用

页码：261 / 269

页数：8

共 23 条

[1] Rehman Z.U., Khan S., Brusseau M.L., Et al., Lead and cadmium contamination and exposure risk assessment via consumption of vegetables grown in agricultural soils of five-selected regions of Pakistan, Chemosphere, 168, pp. 1589-1596, (2017)
[2] Luce M.S., Ziadi N., Gagnon B., Et al., Visible near infrared reflectance spectroscopy prediction of soil heavy metal concentrations in paper mill biosolid- and liming by-product-amended agricultural soils, Geoderma, 288, pp. 23-36, (2017)
[3] Marin A., Andrades M., Inigo V., Et al., Lead and cadmium in soils of La Rioja Vineyards, Spain: pb and cd in Vineyards, Land Degradation and Development, 27, 4, pp. 1286-1294, (2016)
[4] Sun W., Zhang X., Zou B., Et al., Exploring the potential of spectral classification in estimation of soil contaminant elements, Remote Sensing, 9, 6, (2017)
[5] Santan F.B.D., Souza A.M.D., Poppi R.J., Et al., Visible and near infrared spectroscopy coupled to random forest to quantify some soil quality parameters, Spectrochimica Acta Part A: Molecular and Biomolecular Spectroscopy, 191, pp. 454-462, (2018)
[6] Xie X., Pan X., Sun B., Visible and near-infrared diffuse reflectance spectroscopy for prediction of soil properties near a copper smelter, Pedosphere, 22, 3, pp. 351-366, (2012)
[7] Wang J., Tiyip T., Zhang D., Spectral detection of chromium content in desert soil based on fractional differential, Transactions of the Chinese Society for Agricultural Machinery, 48, 5, pp. 152-158, (2017)
[8] Vahid K., Faramarzdoulati A., Saeed Y., Et al., Monitoring soil lead and zinc contents via combination of spectroscopy with extreme learning machine and other data mining methods, Geoderma, 318, pp. 29-41, (2018)
[9] Frank R., Michael D., Ingo M., Et al., Prediction of soil parameters using the spectral range between 350 and 15000 nm: a case study based on the permanent soil monitoring program in Saxony, Germany, Geoderma, 315, pp. 188-198, (2018)
[10] Chen T., Chang Q., Clevers J.G.P.W., Et al., Rapid identification of soil cadmium pollution risk at regional scale based on visible and near-infrared spectroscopy, Environmental Pollution, 206, pp. 217-226, (2015)

← 1 2 3 →