荧光寿命分析法鉴别机油品种特征信息提取

由于机油的品种鉴别在环境监测、事故认定等领域有着重要的作用,结合非线性最小二乘法的荧光寿命衰减曲线提取技术,能解决不同机油混合后的分类识别问题。研究通过激光诱导荧光雷达系统,利用时间分辨荧光实验,采用非线性最小二乘法将荧光寿命衰减曲线按照不同指数拟合,从相关指数及残差分布确定最佳拟合曲线。实验结果表明,三种不同机油产品的荧光寿命衰减曲线按照不同指数衰减其拟合效果不同,选取各自最佳拟合曲线,按照合成规律可以拟合出不同机油的荧光平均寿命参数,通过误差分析可信赖程度在95%以上。当以所有激发区域内像素点的平均荧光寿命为中心,利用2倍标准差作为置信限,统计其中两种不同机油激发区域中像素点荧光寿命值落入各自置信区间内的概率,二者的概率分别为87%和68%。通过三种不同油品的对比试验,验证了利用荧光寿命分析法实现不同机油产品油种识别的可能性,并可通过荧光寿命成像判断不同油种的空间分布信息。荧光寿命分析法为机油产品油种鉴别、成分分析及检测空间分布提供了一种新途径。

Feature information extraction in identification of oil varietiesbased on fluorescence lifetime analysis method

Identification of oil varieties play an important role in environmental monitoring,accident identification and other fields.Based on the complex composition of oil products,the fluorescence lifetime imaging technique combined with nonlinear least square method was proposed to fit the fluorescence life attenuation curve and extract mean fluorescence lifetime parameters,and thus solved the classification and identification problem of different oil mixtures.Using time-resolved fluorescence experiment in laser-induced fluorescence system,the best fitting curve was determined from the correlation index and residual distribution.Experimental results show that the fluorescence life attenuation curves of three different oil products had different fitting effect according to different exponential attenuation curves.The optimal fitting curves were selected to fit the average fluorescence life parameters of different oil products according to the synthesis method,and trustiness of the results was above 95%.Taking the average fluorescence life of all excited pixels as the center,using 2 times standard deviations as the confidence limit,the probability of pixel fluorescence lifetime value falling within their confidence intervals in two different oil excitation regions was calculated as 87% and 68%,respectively.Comparing the mixture of three different oil products,the possibility of using fluorescence life analysis to identify different oil type was verified,and the spatial distribution information of different oil types was determined by fluorescence life imaging.Fluorescence life analysis provides a new approach for oil product identification,component analysis and spatial distribution detection.