Machine learning-based prediction of biocrude yields and higher heating values from hydrothermal liquefaction of wet biomass and wastes: A literature review-derived dataset study

Global warming, primarily driven by greenhouse gas emissions, presents a critical challenge that necessitates innovative solutions. This study explores biomass hydrothermal liquefaction as a promising method to mitigate this issue by converting organic matter into renewable bio-oil. HTL processes biomass, including agricultural and forestry residues, under high temperature and pressure with water, resulting in a high-quality bio-oil compatible with existing fuel infrastructure. Machine learning emerges as a crucial tool in modeling and optimizing the HTL process due to its complex and nonlinear nature. ML algorithms can predict process behavior, optimize operational conditions, and enhance overall efficiency and sustainability. By handling extensive datasets, these algorithms aim to predict product properties and minimize environmental impacts. The dataset compiled in this study is pivotal for advancing ML applications in HTL, providing researchers with valuable information to refine models and reduce data collection time. This comprehensive dataset includes details on the biological properties of biomass, extending the scope of existing studies.由温室气体排放主导的全球变暖，已成为一项亟需创新性解决方案的严峻挑战。本研究探讨了生物质水热液化（biomass hydrothermal liquefaction, HTL）作为缓解该问题的颇具潜力的方法，其可将有机物质转化为可再生生物油。该技术以水为反应介质，在高温高压条件下处理包括农业与林业废弃物在内的各类生物质，可产出与现有燃料基础设施兼容的高品质生物油。由于水热液化过程具有复杂非线性的特性，机器学习（Machine Learning, ML）已成为该过程建模与优化的关键工具。机器学习算法能够预测过程运行行为、优化操作工况，并提升整体效率与可持续性。通过处理大规模数据集，此类算法可实现产物属性预测，并最大限度降低环境负面影响。本研究构建的数据集，对于推动机器学习在水热液化领域的应用至关重要，可为研究人员提供宝贵数据以优化模型，并缩短数据收集周期。该综合数据集涵盖了生物质的生物学属性细节，拓展了现有研究的覆盖范围。