论文部分内容阅读
土壤中石油类含量的检测对石油污染的预防与治理具有重要的实际意义。本文首先进行孤东油田土壤样品高光谱反射率的室内测定及石油含量的检测,然后利用单变量预测模型和逐步回归方法分析了土壤光谱特征参数与石油类含量之间的线性和非线性关系,结果表明:包络线分析的第三折线段斜率与石油类含量相关性最好,该段斜率的三次曲线函数为石油类含量的最佳单变量估算模型。标准正态变量变换对光谱的预处理效果最好,利用变换后光谱建立多元模型,其调整的判定系数R2是0.826,总均方根RMSE是0.531,且自变量个数较少,为最优预测模型。本文提出的利用高光谱数据检测土壤中石油类含量的方法,为土壤石油类污染检测提供了一种有效的新思路。
The detection of oil content in soil has important practical significance for the prevention and treatment of petroleum pollution. In this paper, firstly, the indoor determination of hyperspectral reflectance and the determination of petroleum content in soil samples of Gudong Oilfield were carried out. Then, the linear and nonlinear relationship between soil spectral characteristic parameters and oil content were analyzed by using univariate prediction model and stepwise regression analysis. The results show that the slope of the third fold line of the envelope analysis is the best one to the oil content, and the cubic curve function of the slope is the best single variable estimation model of the petroleum content. The normalized normal variable transform has the best pretreatment effect on the spectrum. The multivariate model was established by using the transformed spectrum. The adjusted coefficient of determination R2 was 0.826, the total root mean square RMSE was 0.531, and the number of independent variables was less, which was the best Predictive model. In this paper, the method of using hyperspectral data to detect oil content in soil is provided, which provides an effective new idea for soil petroleum pollution detection.