基于三维荧光光谱技术结合交替加权残差约束四线性分解的不同盐度条件下混合油液检测

Mixed Oil Detection Based on 3D Fluorescence Spectroscopy Combined with AWRCQLD under Different Salinity Conditions

石油作为一种重要的化石能源,是人类社会生产活动中不可缺少的一部分。石油在被人们开采、使用的过程中不可避免地会发生泄漏,泄漏的石油会给生态环境带来严重的威胁。因此,在石油泄漏后需要及时对其进行处理,而其前提是能够准确识别石油种类。由于石油中多种物质具有荧光特性,因此应用荧光光谱法可对石油进行有效检测。但石油所含组分较多,使得其光谱信息重叠严重,识别困难。而三阶校正方法具有"三阶优势",可以分辨高共线性、高噪声水平下的数据。其中,三阶校正中的交替加权残差约束四线性分解(AWRCQLD)算法具有收敛速度快、对组分数不敏感等优点;因此,利用三维荧光光谱技术结合AWRCQLD算法,对混合油液进行检测。首先,配制3种盐度条件下的十二烷基硫酸钠(SDS)溶剂;并在每种盐度条件下分别将航空煤油和润滑油按照不同浓度比混合,最终得到24个校正样本和9个预测样本。然后,使用FLS920荧光光谱仪对实验样本进行光谱数据采集。其次,使用扣除空白法去除光谱中的散射,并通过核一致诊断法判断混合油中的组分数。最后,用AWRCQLD算法对四维光谱矩阵进行解析。研究结果表明,在0~20盐度范围内,随着盐度的增加,航空煤油的荧光强度先减小后增大,润滑油的荧光强度先增大后减小;混合油解析光谱曲线分别与航空煤油及润滑油的实际光谱曲线重合度良好;经AWRCQLD算法解析后得到的航空煤油的回收率范围为100.2%~109%,均方根误差为0.0021 mg·mL^-1;润滑油的回收率范围为91.8%~109.3%,均方根误差为0.0048 mg·mL^-1。通过引入盐度作为新一维度的数据,从而将三维光谱数据阵扩展到相应的四维光谱数据阵。并利用AWRCQLD算法对四维光谱数据阵进行了解析,实现了在不同盐度条件下对混合油的定性和定量分析。同时,为不同盐度条件下的混合油液检测提供了参考。

As an important fossil energy source,oil is an indispensable part of human society’s production activities.When the oil is mined and used,it could be leaked inevitably.The leaked oil will pollute the ecological environment.Therefore,it is necessary to deal with oil spills in a timely manner.Accurate identification of petroleum species is a prerequisite for handling oil spills.Petroleum contains a variety of substances with fluorescent properties.Therefore,fluorescence spectroscopy is an effective method for detecting petroleum.Due to a large number of components in the oil,the spectral information overlaps seriously,and the identification is difficult.The third-order calibration method has the"third-order advantage".It can distinguish the data under high collinearity and high noise level.Alternating weighted residue constraint quadrilinear decomposition(AWRCQLD)algorithm is a third-order correction method.AWRCQLD algorithm has the advantages of faster convergence speed and insensitivity to component numbers.Therefore,in this paper,the three-dimensional(3 D)fluorescence spectroscopy combined with AWRCQLD algorithm is used to detect the mixed oil.First,sodium dodecyl sulfate(SDS)was prepared as a solvent under three salinity conditions.Under each salinity condition,jet fuel and lube were mixed according to different concentration ratios.Thus,24 calibration samples and 9 prediction samples are obtained.Secondly,using FLS920 fluorescence spectrometer to acquire spectral data of the experimental samples.Then,the effect of scattering was removed by using blank subtraction,and the number of components in the mixed oil is estimated by the core consistent diagnosis method.Finally,using the AWRCQLD algorithm to analysis the four-dimensional spectral matrix.The results show that,in the range of 0~20 salinity,the fluorescence intensity of jet fuel decreases first and then increases,but the fluorescence intensity of lube increases first and then decreases.The analytical spectral curves of the mixed oils are in good agreement with the actual spectral curves of the jet fuel and lube.The recovery rate of jet fuel obtained by AWRCQLD algorithm is 100.2%~109%and the root mean square error is 0.0021 mg·mL^-1;the recovery rate of lube is 91.8%~109.3%and the root mean square error is 0.0048 mg·mL^-1.By introducing the salinity of seawater as a new dimension of data,the three-dimensional spectral data array is superimposed on this dimension to obtain the four-dimensional spectral data array.In this paper,the four-dimensional spectral data matrix is analyzed by the AWRCQLD algorithm.The purpose of qualitative and quantitative analysis of mixed oil under different salinity conditions is achieved.At the same time,this paper provides a reference for detecting petroleum mixed oil under different salinity conditions.