基于分数阶微分的荒漠土壤铬含量高光谱检测

发布时间：2018-04-28 22:27

本文选题：荒漠土壤 + 重金属　；参考：《农业机械学报》2017年05期

【摘要】：为解决高光谱检测土壤中痕量级重金属含量存在的困难,提高土壤重金属铬含量检测的准确度,利用新疆准东煤田周边168个荒漠土壤样本的重金属铬含量及其对应的高光谱数据,运用分数阶微分算法进行光谱数据预处理,最后利用全部波段进行偏最小二乘建模并进行可视化分析,旨在探讨分数阶微分预处理在高光谱数据估算荒漠土壤重金属铬含量的可能性。结果表明:原始光谱与吸光率变换的分数阶微分模型均在1.8阶微分处达到了最好的精度效果。吸光率变换1.8阶微分模型为最优模型,模型的校正均方根误差为7.68 mg/kg,R_c~2=0.83,预测均方根误差为8.39 mg/kg,R_p~2=0.78,相对分析误差为2.14。最后利用铬含量实测值与光谱预测值通过反距离加权法插值获得研究区土壤重金属铬含量的空间分布,说明利用该方法对土壤重金属铬含量定量检测并进行大尺度的空间分布反演在一定程度上是可行的,为荒漠土壤重金属污染状况的高光谱检测提供了一定的科学依据和技术支持。
[Abstract]:In order to solve the difficulty of detecting trace heavy metals in soil by hyperspectral method, and to improve the accuracy of the determination of chromium in soil, Based on the heavy metal chromium content and its corresponding hyperspectral data of 168 desert soil samples around Zhendong coalfield in Xinjiang, the fractional differential algorithm was used to preprocess the spectral data. Finally, the partial least square modeling and visualization analysis were carried out in all bands to explore the possibility of fractional-order differential pretreatment in estimating the content of heavy metal chromium in desert soil by hyperspectral data. The results show that the fractional-order differential model of the original spectrum and absorptivity transformation achieves the best accuracy at the 1.8 order differential. The 1.8-order differential model of absorptivity transformation is the best model. The corrected RMS error of the model is 7.68 mg / kg ~ (-1) 路kg ~ (-1) ~ (-1). The RMS error of prediction is 8.39 mg / kg ~ (-1) 路kg ~ (-1) ~ 0.78, and the relative analysis error is 2.14 mg 路kg ~ (-1) 路kg ~ (-1) ~ (-1) ~ (-1). Finally, the spatial distribution of heavy metal chromium content in the soil of the study area was obtained by using the measured value of chromium content and the predicted value of spectrum through the inverse distance weighting method. It shows that it is feasible to use this method to quantitatively detect the content of heavy metal chromium in soil and to invert the spatial distribution of the heavy metal on a large scale to a certain extent. It provides scientific basis and technical support for hyperspectral detection of heavy metal pollution in desert soil.
【作者单位】：新疆大学资源与环境科学学院;新疆大学绿洲生态教育部重点实验室;
【基金】：“十二五”国家科技支撑计划项目(2014BAC15B01) 国家自然科学基金重点项目(41130531)
【分类号】：X833

【相似文献】