Abstract Near infrared spectroscopy combined with multivariate calibration such as partial least squares regression is a promising technique for on-line monitoring of anaerobic digesters. Different substrates are used in digesters, depending on their availability and their methanogen potential, to optimize the process. In Europe, the feedstock for anaerobic digesters is dominated by slurry and food waste which are respectively highly biodegradable and fat-containing substrates. The monitoring of the anaerobic digestion process based on digestates coming from these substrates presents some difficulties. The digestion of highly biodegradable substrates comes with the presence of water, which hinders spectroscopic calibration. And fat-containing substrates could lead to the accumulation of long chain fatty acids which are quite difficult to detect in the infrared region. While all existing studies have explored adapted spectroscopic measurements to improve the process monitoring, this study investigated the use of NIRS combined with multi-block analysis to track important anaerobic digestion stability parameters. Infrared measurements can come from several sources in the process monitoring. In addition, sequential and orthogonalized partial least squares have proven their ability of exploiting the underlying relation between several data blocks. These multi-block methods are powerful chemometric tools which can be applied in the monitoring of anaerobic digestion. Polarization light spectroscopy which is also known to improve the comprehension of scattering media like the digestate was also studied.

中文翻译：

---

基于红外光谱的多块SO-PLS方法用于厌氧消化过程监测

摘要 近红外光谱与偏最小二乘回归等多元校准相结合是一种很有前景的厌氧消化器在线监测技术。消化器中使用不同的底物，这取决于它们的可用性和产甲烷的潜力，以优化过程。在欧洲，厌氧消化器的原料主要是泥浆和食物垃圾，它们分别是高度可生物降解和含脂肪的底物。基于来自这些底物的消化物的厌氧消化过程的监测存在一些困难。高度可生物降解的底物的消化伴随着水的存在，这会阻碍光谱校准。含脂肪的底物会导致长链脂肪酸的积累，而这些长链脂肪酸在红外区域很难检测到。虽然所有现有研究都探索了适应光谱测量以改进过程监测，但本研究调查了 NIRS 结合多块分析的使用，以跟踪重要的厌氧消化稳定性参数。红外测量可以来自过程监控中的多个来源。此外，顺序和正交化的偏最小二乘法已经证明了它们利用几个数据块之间的潜在关系的能力。这些多块方法是强大的化学计量工具，可用于监测厌氧消化。