群智能优化的动态基线调整在光纤SPR传感器信号处理中的应用  

Application of Dynamic Baseline Adjustment Based on Swarm Intelligence Optimization in the Signal Processing of FIBER SPR SEnsor

作　　者：戴峻峰 付丽辉 DAI Junfeng;FU Lihui(Faculty of Electronic Information Engineering,Huaiyin Institute of Technology,Huai’an Jiangsu 223003,China;Faculty of Automation,Huaiyin Institute of Technology,Huai’an Jiangsu 223003,China)

机构地区：[1]淮阴工学院电子信息工程学院,江苏淮安223003 [2]淮阴工学院自动化学院,江苏淮安223003

出　　处：《传感技术学报》2023年第7期1092-1102,共11页Chinese Journal of Sensors and Actuators

基　　金：国家自然科学基金项目(62141502);江苏省建设系统科技项目(2019ZD0064);淮安市“1111”工程合作项目(Z413H22507)。

摘　　要：将谱宽(β)、基线下测量点数(m)、面积比(λ)作为最佳动态基线的选择依据,在两级适应度函数条件下,利用粒子群优化算法跟踪β、m、λ,寻找最佳组合参数下的动态基线,有效克服了传统基线选择较为随意的问题。搭建了光纤SPR测试装置,采集8种不同浓度的蔗糖标准溶液的光谱数据,将4种质心跟踪法进行实验对比。结果表明群智能优化动态基线法的预测精度和相关性均最好,可以有效提取质心参数,消除光源波动对测试结果的影响,为实现光纤SPR实时在线监测提供了一种有效途径,其共振波长拟合度为0.996 3,RMSE为1.78。另外,考察了三种质心计算参数的主导地位,结果表明:在设定相同适应度情况下,固定m方法效果最好,进而为质心参数的选择控制提供有效依据。The spectral width(β),number of measurement points below the baseline(m),area ratio(λ)are taken as the basis for selec-ting the best dynamic baseline.Under the condition of two-level fitness function,particle swarm optimization algorithm is used to track and find the dynamic baseline under the best combination of parameters,which effectively overcomes the problem of random selection of baseline in traditional method.The spectral data of standard sucrose solutions of 8 different concentrations are collected by building the fiber SPR test device,and four centroid methods are compared experimentally.The results show that the swarm intelligence optimization dynamic baseline method has the best prediction accuracy and correlation,it can effectively extract the centroid parameters and eliminate the influence of light source fluctuation,which provides an effective way to realize the real-time online monitoring of fiber SPR,its fitting degree of resonance wavelength is 0.9963 and RMSE is 1.78.In addition,the dominant position of the three parameters is investigated,the results show that when the same fitness is set,the fixed m method has the best effect,which provides an effective ba-sis for the selection of centroid parameters.

关 键 词：光纤SPR传感器 表面等离子共振效应 粒子群算法 群智能优化 动态基线 

分 类 号：TP212.3[自动化与计算机技术—检测技术与自动化装置]