上海农业学报2025，Vol.41Issue(6)：32-39,8.DOI:10.15955/j.issn1000-3924.2025.06.05
双孢蘑菇三次料关键品质近红外光谱预测模型的构建
Construction of near-infrared spectroscopy prediction model for key quality of Agaricus bisporus compost
摘要
Abstract
In this study,a total of 272 Phase Ⅲ compost samples were collected,and near-infrared(NIR)spectra with wavelengths ranging from 850 nm to 2 498 nm were acquired using a NIR spectrometer.The spectral data were preprocessed using WinISI Ⅲ software,and the prediction models for nitrogen content,ash content and moisture content were developed using Modified Partial Least Squares(MPLS)regression.The optimal models were selected based on the standard error of cross validation(SEC V)and 1 minus the variance ratio(1-VR).Additionally,60 samples were used to evaluate the predictive performance of the models.The results demonstrated that the best NIR prediction model of nitrogen content was achieved using None+first derivative with SECV of 0.083 9 and 1-VR of 0.806 0.The optimal model for ash content was obtained with standard normalization+first derivative,producing an SECV of 1.899 1 and a 1-VR of 0.785 1.For moisture content,the highest-performing model utilized Detrend+first derivative,with an SECV of 1.290 2 and a 1-VR of 0.779 7.Validation using the 60 independent samples confirmed that the measured values of nitrogen,ash,and moisture contents showed no significant differences from the predicted values,indicating that the model had good predictive performance.The NIR models established in this study have high accuracy and good stability,demonstrating their potential for rapid and reliable assessment of key composting quality parameters.关键词
双孢蘑菇/三次料/近红外光谱/改进最小二乘法Key words
Agaricus bisporus/Phase Ⅲ compost/Near-infrared spectroscopy/Modified Partial Least Squares分类
农业科技引用本文复制引用
肖婷婷,王倩,陈万超,张津京,张宇辰,汤倩倩,陈辉..双孢蘑菇三次料关键品质近红外光谱预测模型的构建[J].上海农业学报,2025,41(6):32-39,8.基金项目
上海市现代农业产业技术体系[沪农科产字(2025)第9号] （2025）
上海市农业科学院AI赋能农业专项课题[沪农科AI-B(2025)004] （2025）
