基于改进鲸鱼算法的波长调制光谱拟合技术  

Study on Wavelength Modulation Spectroscopy(WMS)Fitting Technique Based on Improved Whale Optimization Algorithm(WOA)

作　　者：王晨茜 段玉杰 朱存光 Wang Chenxi;Duan Yujie;Zhu Cunguang(School of Physics Science and Information Technology,Liaocheng University,Liaocheng 252000)

机构地区：[1]聊城大学物理科学与信息工程学院,山东聊城252000

出　　处：《中阿科技论坛(中英文)》2025年第3期81-85,共5页China-Arab States Science and Technology Forum

基　　金：国家自然科学基金“基于改进鲸鱼算法的波长调制光谱拟合技术”(61705080)。

摘　　要：为提高光谱特征提取的精度和效率,文章提出了一种基于改进鲸鱼优化算法(WOA)的波长调制光谱(WMS)拟合技术。通过引入非线性时变自适应权重和差分变异微扰因子,对传统鲸鱼优化算法进行优化,使其全局搜索能力和收敛速度得到显著提升。以乙炔在1 532.82 nm处的P13吸收线为例,文章对比了改进鲸鱼算法、传统鲸鱼算法以及Levenberg-Marquardt(LM)算法在气体体积分数和激光调谐参数检索中的性能。研究结果表明,改进后的鲸鱼算法能够稳定且高效地提取WMS-2 f/1 f光谱特征,拟合精度更高,残差可降至10-5数量级,拟合时间缩短至20 s,自由参数的相对误差均控制在0.3%以内,显著优于其他拟合算法。In order to enhance the accuracy and efficiency of spectroscopy feature extraction,a WMS fitting technique based on improved WOA was proposed.By introducing non-linear time-varying adaptive weights and differential variation perturbation factors,the traditional WOA is optimized to significantly boost its global search capability and convergence speed.Taking the P13 absorption line of acetylene at 1532.82 nm as an example,this article compares the performance of the improved WOA,traditional WOA,and Levenberg Marquardt(LM)algorithms in gas volume fraction and laser tuning parameter retrieval.The results suggest that the improved WOA is significantly better than other algorithms in that it can efficiently extract WMS-2 f/1 f spectroscopy features,with higher accuracy and shorter time(20 seconds),and the residual can be reduced to the order magnitude of 10-5,the relative error of free parameters is controlled within 0.3%.

关 键 词：波长调制光谱 光谱拟合 鲸鱼优化算法 

分 类 号：TN929.11[电子电信—通信与信息系统]