Abstract:

With the annual increase in the output of organic solid waste, how to achieve its efficient treatment and resource utilization has become a research hotspot in the current environmental field. High-solids anaerobic digestion is regarded as the core technology to achieve the resourcefulness of organic solid waste, but faces problems such as system instability triggered by high solids loading and low gas production efficiency. Conductive materials are widely used in the field of high-solids anaerobic digestion of organic solid waste by virtue of their conductivity, pore structure and redox activity and other characteristics of regulating microbial metabolism and electron transfer. On the basis of existing studies, the enhanced effects of iron-based, carbon-based and iron-carbon composites on high solid anaerobic digestion were summarized, and the specific mechanisms behind the performance enhancement of the system by conductive materials were elucidated based on three perspectives: modulation of key enzyme activities, optimization of functional microbial communities, and enhancement of direct interspecies electron transfer. The potential application of machine learning models in the prediction of methane production efficiency and identification of key parameters is further explored, providing new ideas for the optimization of conductive material parameters. Meanwhile, the future research direction of conducting materials for enhanced high solid anaerobic digestion of organic solid wastes is also prospected.

Key words: conductive materials, high-solids anaerobic digestion, microbial communities, direct interspecies electron transfer, machine learning

摘要：

随着有机固体废物产量的逐年递增，如何实现其高效处理与资源化已成为当下环境领域的研究热点。高固厌氧消化被视为实现有机固体废物资源化的核心技术，但面临着高固体负荷引发的系统失稳及产气效率低等问题。导电材料凭借自身导电性、孔隙结构及氧化还原活性等特性调控微生物代谢与电子传递，被广泛应用于有机固废的高固厌氧消化处理领域。在现有研究基础之上，总结了铁基、碳基及铁碳复合材料对高固厌氧消化的强化效应。从关键酶活性调控、功能微生物群落优化和直接种间电子传递强化三个角度，阐明了导电材料提升系统性能背后的具体作用机制。进一步探讨了机器学习模型通过产甲烷效能预测与关键参数识别在导电材料参数优化中的应用潜力。同时，对导电材料强化有机固废高固厌氧消化的未来研究方向进行了展望。

关键词: 导电材料, 高固厌氧消化, 微生物群落, 直接种间电子转移, 机器学习