期刊
ENERGY
卷 243, 期 -, 页码 -出版社
PERGAMON-ELSEVIER SCIENCE LTD
DOI: 10.1016/j.energy.2021.122783
关键词
Fine screenings; Combustion; Thermo-kinetic; Gas products; Deep learning
资金
- National Key R&D Program of China [2019YFC1904001]
- National Natural Science Foundations of China [52000143]
This study explored the combustion behavior, kinetics, thermodynamics and gas products of fine screenings from municipal solid waste in an air atmosphere. A deep learning model using the 1D-CNN-LSTM algorithm was established to predict the thermogravimetric data of combustion, with visualization technology applied to display the weights and biases. The study revealed the stages of combustion and the changes in activation energy at different stages. The highest enthalpy and free Gibbs energy were observed in stage IV, while the highest changes in entropy occurred in stage II. Gas products and functional groups were also detected. The 1D-CNN-LSTM model achieved excellent prediction performance with optimal settings.
The combustion behavior, kinetics, thermodynamics and gas products of fine screenings (FS) classified from municipal solid waste (MSW) in an air atmosphere were explored by TG-FTIR. A deep learning model was established using 1D-CNN-LSTM algorithm to predict thermogravimetric data of FS com-bustion, with visualization technology (TensorBoard) applied to display the weights and biases in various cells. The thermogravimetric analysis (TG) and differential thermal gravity (DTG) curves indicated that the FS combustion process can be divided into four stages. The average activation energy (E-a) of FS combusted at different stages, exhibited different change tendencies with increasing levels of conversion (alpha). The highest enthalpy (delta H) of 206.40 kJ/mol and free Gibbs energy (delta G) of 55.03 kJ/mol emerged in stage IV, while the highest changes of entropy (delta S) of 169.11 J/(mol.K) occurred in stage II. The main gas products (CO2, H2O and CO) and functional groups (C=& nbsp;O and phenols) were all detected. For the 1D-CNN-LSTM model, the optimal settings for the prediction of thermogravimetric data were a neuron number of 150, dropout of 0.003, epoch number of 200, and batch size of 25. The highest correlation coefficient (R-2) of 94.41% was obtained using the optimum model parameters, achieving an excellent prediction performance. (C)& nbsp;2021 Elsevier Ltd. All rights reserved.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据