论文标题
挥发性脂肪酸的动态优化，使用深度学习神经网络富集生物氢化
Dynamic optimization of volatile fatty acids to enrich biohydrogen production using a deep learning neural network
论文作者
论文摘要
制定了一种新的策略来研究挥发性脂肪酸（VFA）对沼气生产效率的影响，重点是改善生物h $ _2 $。从UASB反应堆处理的家禽屠宰场废水中获得的接种物，厌氧颗粒状污泥已通过五种不同的预处理进行了预处理。 VFA和沼气化合物之间的关系作为时间依赖性成分。在具有较小样本量数据的时间依赖性过程中，回归模型在估计响应方面可能还不够好。因此，开发了一个深度学习神经网络（DNN）模型，以估计基于VFA的沼气化合物。该模型预测沼气化合物的准确性高于多元回归模型。此外，它可以预测时间变化对沼气化合物的影响。分析表明，所有预处理能够成功地增加丁酸 /乙酸的比率，大幅度降低丙酸，并提高生物h $ _2 $产生的效率。正如发现的那样，丁酸对生物H $ _2 $的影响最大，而丙酸对CH $ _4 $的产生的影响最大。使用优化方法，集成的DNN和可取性分析确定了最佳数量的VFA，并根据消化时间动态重新训练。因此，乙酸，丙酸和丁酸的最佳范围分别为823.2-1534.3、36.3-47.4和1522-1822 mg/l，分别确定为25.23-1233.63 h的消化时间。这些值导致生产生物h $ _2 $，n $ _2 $，co $ _2 $和ch $ _4 $ in范围为6.4-26.2、12.2-43.2-43.2、5-25.3和0-1.4 mmol/l。 VFA的最佳范围是相对较宽的范围，实际上可以在沼气植物中使用。
A new strategy was developed to investigate the effect of volatile fatty acids (VFAs) on the efficiency of biogas production with a focus on improving bio-H$_2$. The inoculum used, anaerobic granular sludge obtained from a UASB reactor treating poultry slaughterhouse wastewater, was pretreated with five different pretreatments. The relationship between VFAs and biogas compounds was studied as time-dependent components. In time-dependent processes with small sample size data, regression models may not be good enough at estimating responses. Therefore, a deep learning neural network (DNN) model was developed to estimate the biogas compounds based on the VFAs. The accuracy of this model to predict the biogas compounds was higher than that of multivariate regression models. Further, it could predict the effect of time changes on biogas compounds. Analysis showed that all the pretreatments were able to increase the ratio of butyric acid / acetic acid successfully, decrease propionic acid drastically, and increase the efficiency of bio-H$_2$ production. As discovered, butyric acid had the greatest effect on bio-H$_2$, and propionic acid had the greatest effect on CH$_4$ production. The best amounts of the VFAs were determined using an optimization method, integrated DNN and desirability analysis, dynamically retrained based on digestion time. Accordingly, optimal ranges of acetic, propionic, and butyric acids were 823.2 - 1534.3, 36.3 - 47.4, and 1522 - 1822 mg/L, respectively, determined for digestion time of 25.23 - 123.63 h. These values resulted in the production of bio-H$_2$, N$_2$, CO$_2$, and CH$_4$ in ranges of 6.4 - 26.2, 12.2 - 43.2, 5 - 25.3, and 0 - 1.4 mmol/L, respectively. The optimum ranges of VFAs are relatively wide ranges and practically can be used in biogas plants.
