基于SPA算法的藤茶风味啤酒近红外光谱波长变量优选

打开文本图片集

关键词：SPA算法；藤茶风味啤酒；近红外光谱；波长变量；特征优选

中图分类号：TP312 文献标识码：A

文章编号：0439-8114（2026）02-0183-05

DOI：10.14088/j.cnki.issn0439-8114.2026.02.027 开放科学（资源服务）标识码（OSID）：

Optimization of near-infrared spectral wavelength variables for vine tea-flavored beer based on the SPA algorithm

GUO Yi-feng

（EnshiPolytechnic，Enshi 44500o，Hubei，China）

Abstract：Thewavelengthvariableselectionmethodforinfraredspectroscopyprimarilyemployedmultipleiterationstoscreenwavelengthvariableswithsignificantmodelcontributionfromthefullspectrumdata.Thisapproachexhibitedpoorresistancetodudnt informationintefrene，makingitdiulttoccuatelyflecthetuhaacteristiofteple.Todssthis，veength variableselectionmethodbasedotheSPAalgorithasproposedfornarfraredspectrosopyofvinetea-flavredbee.ultiscat teringcorretionwasappliedtoteear-ifraredspetraldatatoeanceabsortioniformationcoreatedwithcomponntcontent andreducenoise.Thecontinuousprojectionalgorithmwasusedforpreliminarywavelengthsrening.Combinedwithcorrelationand colinearitypenaltisfortheargetcomponentsinwsteriateaflavrdetectionisapproachselectedawavelengthsubsewithlownformationredundancyandstrong predictivecapability.Combinedentropyandmutual informationwereusedtoasess wavelengthedundancy，constructingascoringfunctionincorporatinginformationcontentandredundancypenalties toselectthesubsetwiththe highestoverallsor.Tesultssowedthataferwavelengtharableoptimzationusingthisthodpectralabsorptioncoveger mained stable at approximately 95% ，indicating highly satisfactory optimization outcomes.

：y Words：SPAalgorithm；vine tea-flavored beer；near-infrared spectroscopy；wavelength variables；feature selecti

藤茶风味啤酒的酿造过程涉及复杂的成分相互作用，传统化学分析方法在快速检测过程中存在效率低下和成本较高的问题。（剩余5780字）