Scholarlink

<p>This study investigates the intensification of anaerobic digestion (AD) through the combined use of phosphogypsum as a mineral additive and advanced pretreatment techniques, including ultrasonic cavitation and electrofermentation. Special focus is placed on modeling these processes using artificial neural networks (ANNs) to address key research gaps such as limited integration of pretreatment effects, the role of complex additives, and insufficient Modeling of biogas production. The experimental setup involved poultry manure digested with phosphogypsum additive, processed in a microbial electrolysis cell and an ultrasonic cavitator, integrated with a standard anaerobic bioreactor. Data collected were analyzed using STATISTICA ANNs to predict biogas yield and process dynamics. Electrofermentation combined with phosphogypsum reduced hydrogen sulfide levels from over 5000&#xa0;mg/L to 487&#xa0;mg/L on day 14 and 97&#xa0;mg/L by day 26. Ultrasonic pretreatment enhanced biogas output by 1.3 times, reaching 9800 mL of total biogas and 4980 mL of methane by day 25. Pretreatment altered the physicochemical properties of substrate, increasing the bioavailability of nutrients and enhancing microbial activity. Model accuracy ranged from 78.2 to 97.5%, confirming the ANN ability to generalize under variable conditions. This study contributes to development of robust, data-driven models that integrate pretreatment strategies and additive effects, providing a more holistic understanding of AD. Findings highlight the potential of combining phosphogypsum, pretreatment methods, and artificial intelligence to optimize biogas production and improve the environmental value of the digestate. The integrated ANN modeling approach provides a valuable tool for process prediction, supporting the development of more sustainable AD systems.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Anaerobic Digestion of Poultry Manure with Phosphogypsum Using Microbial Electrolysis and Ultrasound: A Neural Network Approach

  • 期刊论文

  • 发表日期：

    2025年9月15日

查看全文
  • Yelizaveta Chernysh,
  • Viktoriia Chubur,
  • Vladimir Shtepa,
  • Dmytro Alekseevsky,
  • Yevhen Bataltsev,
  • Hynek Roubík

摘要

This study investigates the intensification of anaerobic digestion (AD) through the combined use of phosphogypsum as a mineral additive and advanced pretreatment techniques, including ultrasonic cavitation and electrofermentation. Special focus is placed on modeling these processes using artificial neural networks (ANNs) to address key research gaps such as limited integration of pretreatment effects, the role of complex additives, and insufficient Modeling of biogas production. The experimental setup involved poultry manure digested with phosphogypsum additive, processed in a microbial electrolysis cell and an ultrasonic cavitator, integrated with a standard anaerobic bioreactor. Data collected were analyzed using STATISTICA ANNs to predict biogas yield and process dynamics. Electrofermentation combined with phosphogypsum reduced hydrogen sulfide levels from over 5000 mg/L to 487 mg/L on day 14 and 97 mg/L by day 26. Ultrasonic pretreatment enhanced biogas output by 1.3 times, reaching 9800 mL of total biogas and 4980 mL of methane by day 25. Pretreatment altered the physicochemical properties of substrate, increasing the bioavailability of nutrients and enhancing microbial activity. Model accuracy ranged from 78.2 to 97.5%, confirming the ANN ability to generalize under variable conditions. This study contributes to development of robust, data-driven models that integrate pretreatment strategies and additive effects, providing a more holistic understanding of AD. Findings highlight the potential of combining phosphogypsum, pretreatment methods, and artificial intelligence to optimize biogas production and improve the environmental value of the digestate. The integrated ANN modeling approach provides a valuable tool for process prediction, supporting the development of more sustainable AD systems.