基于近红外光谱技术的豆腐渣堆肥有机质含量测定  

Determination of the organic matter content of tofu residue compost based on near⁃infrared spectroscopy

作　　者：刘晓吉 苏云 徐佳佳 毛勇位 李卫华 薛同站 孙岩松 周阳 LIU XiaoJi;SU Yun;XU JiaJia;MAO YongWei;LI WeiHua;XUE TongZhan;SUN YanSong;ZHOU Yang(China Energy Conservation(Feixi)Environmental Protection Energy Co.,Ltd.,Hefei 231241;School of Environment and Energy Engineering,Anhui Jianzhu University,Hefei 230601;School of Architecture and Engineering,Tongling College,Tongling 244061;China Environmental Protection Group Co.,Ltd.,Beijing 100082,China)

机构地区：[1]中节能(肥西)环保能源有限公司,合肥231241 [2]安徽建筑大学环境与能源工程学院,合肥230601 [3]铜陵学院建筑工程学院,铜陵244061 [4]中国环境保护集团有限公司,北京100082

出　　处：《北京化工大学学报（自然科学版）》2024年第5期20-28,共9页Journal of Beijing University of Chemical Technology(Natural Science Edition)

基　　金：中国节能环保集团有限公司重大科技创新项目(cecep-zdkj-2020-001);国家重点研发计划(2020YFC1908603);国家自然科学基金(51978003);安徽高校自然科学研究项目(2022AH050258);环境污染控制与废弃物资源化利用教育厅创新研究团队项目(2022AH010019)。

摘　　要：为了对堆肥过程中有机质含量进行实时动态分析,以豆腐渣作为底物辅以完全发酵的厨余垃圾粉末进行混合好氧堆肥,通过近红外光谱分析技术采集不同处理阶段堆肥样品的光谱数据,分别采用归一化法、一阶微分法和二阶微分法对原始光谱进行预处理,利用反向区间偏最小二乘法(biPLS)、联合区间偏最小二乘法(siPLS)和间隔区间偏最小二乘法(iPLS)构建近红外光谱吸光度与有机质含量之间的定量分析模型。结果显示,采用二阶微分预处理方法结合iPLS建立的模型性能最佳,第23个子区间对应的最佳特征波段为5 832~6 086 cm-1,校正集的相关系数(Rc)为0.986 1,交互验证均方根误差(RMSECV)为0.824 7,预测集的相关系数(Rp)为0.964 7,均方根误差(RMSEP)为0.445 7,相对分析误差(RPD)为3.8。结果表明所建立的模型具有良好的稳定性和可靠性,二阶微分预处理方法结合iPLS可以有效地优化光谱建模区域,提高模型预测能力,实现堆肥样品有机质含量的快速测定。In order to dynamically analyze the content of organic matter during the composting process in real time,tofu residue was used as the substrate supplemented with fully fermented kitchen waste powder for mixed aerobic composting.The near-infrared spectral data for composting samples were collected at different treatment stages.The original spectra were pretreated by a normalization method,a first-order differential method and a second-order differential method.Backward interval partial least squares(biPLS),synergy interval partial least squares(siPLS)and interval partial least squares(iPLS)were used to construct quantitative analysis models of the near-infrared spectral absorbance and organic matter content.The results show that the model established using the second-order differential pretreatment method combined with iPLS has the best performance.The best characteris-tic band corresponding to the 23rd sub-interval was 5832-6086 cm-1,the correlation coefficient of the calibration set(Rc)was 0.9861,the root mean square error of cross-validation(RMSECV)was 0.8247,the correlation coef-ficient of the prediction set(Rp)was 0.9647,the root mean square error(RMSEP)was 0.4457,and the relative predictive deviation(RPD)was 3.8.The results show that the established model has good stability and reliability.The second-order differential pretreatment method combined with iPLS can effectively optimize the spectral model-ing area,improve the prediction ability of the model,and realize the rapid determination of the organic matter con-tent in compost samples.

关 键 词：豆腐渣 堆肥有机质 近红外光谱 偏最小二乘法 定量分析模型 

分 类 号：X705[环境科学与工程—环境工程]