Scholarlink

<p>This work deals with the modeling of the enzymatic hydrolysis of pretreated sugarcane bagasse (SB) for fermentable sugars production, where the response surface methodology (RSM) and the adaptive neuro-fuzzy inference system (ANFIS) approach were evaluated. Fuzzy logic is one of the many techniques used by artificial intelligence, which seeks to create intelligent systems capable of solving complex problems and learning from available information. Enzymatic hydrolysis (pH 5.0) of pretreated SB was performed at laboratory (bottles) using commercial cellulase (Sigma, obtained from <i>A. niger</i> with activity of 1.47 U.mg<sup>− 1</sup>) in a shaker incubator with 120&#xa0;rpm and 50&#xa0;°C. Initially, the RSM was used for evaluating the effects of three variables of hydrolysis and subsequently, ANFIS was tested. The input variables considered in the models were hydrolysis time (t), enzyme concentration (E), and substrate concentration (S), while the yield of sugars (glucose) served as the response (output) variable. The RSM modeling showed a good fitting in this work (R<sup>2</sup> = 0.9859). The ANFIS tool efficiently predicted the glucose yield (R<sup>2</sup> = 0.9992). The optimal response, achieving a glucose yield of 25.0&#xa0;g L<sup>− 1</sup> occurred at process settings of t = 60&#xa0;h, E = 3.3%, and S = 23.3&#xa0;g L<sup>− 1</sup>. In conclusion, the ANFIS methodology represents an interesting alternative for modeling of complex chemical processes, especially in those cases where RSM falls short in achieving satisfactory results in terms of model fitting.</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

Modeling of Enzymatic Hydrolysis of Sugarcane Bagasse for Fermentable Sugar Production Using Response Surface Methodology and Adaptive Neuro-fuzzy Inference System

  • 期刊论文

  • 发表日期：

    2025年11月6日

查看全文
  • Debora Guerino Boico,
  • Salah Din Mahmud Hasan,
  • João Vitor Pessini,
  • Jéssyca Ketterine Carvalho,
  • Edson Antonio da Silva,
  • Emmanuel Zullo Godinho,
  • Fernando de Lima Caneppele

摘要

This work deals with the modeling of the enzymatic hydrolysis of pretreated sugarcane bagasse (SB) for fermentable sugars production, where the response surface methodology (RSM) and the adaptive neuro-fuzzy inference system (ANFIS) approach were evaluated. Fuzzy logic is one of the many techniques used by artificial intelligence, which seeks to create intelligent systems capable of solving complex problems and learning from available information. Enzymatic hydrolysis (pH 5.0) of pretreated SB was performed at laboratory (bottles) using commercial cellulase (Sigma, obtained from A. niger with activity of 1.47 U.mg− 1) in a shaker incubator with 120 rpm and 50 °C. Initially, the RSM was used for evaluating the effects of three variables of hydrolysis and subsequently, ANFIS was tested. The input variables considered in the models were hydrolysis time (t), enzyme concentration (E), and substrate concentration (S), while the yield of sugars (glucose) served as the response (output) variable. The RSM modeling showed a good fitting in this work (R2 = 0.9859). The ANFIS tool efficiently predicted the glucose yield (R2 = 0.9992). The optimal response, achieving a glucose yield of 25.0 g L− 1 occurred at process settings of t = 60 h, E = 3.3%, and S = 23.3 g L− 1. In conclusion, the ANFIS methodology represents an interesting alternative for modeling of complex chemical processes, especially in those cases where RSM falls short in achieving satisfactory results in terms of model fitting.