1. Prediction of MSW pyrolysis products based on a deep artificial neural network.
- Author
-
Zang Y, Ge S, Lin Y, Yin L, and Chen D
- Subjects
- Temperature, Neural Networks, Computer, Biomass, Pyrolysis, Solid Waste analysis
- Abstract
Pyrolysis is a promising method for recovering resources and energy products from municipal solid waste (MSW). Predicting MSW pyrolysis products is crucial for establishing an efficient pyrolysis system for resource recovery. In this study, a database was established based on MySQL to record relevant information on MSW pyrolysis, which includes the MSW ultimate analysis results, proximate analysis results, parameters of pyrolysis operation and yields of pyrolysis products, etc. Based on the database and with help of a deep artificial neural network (ANN) which contains 10 hidden layers, a prediction model was successfully established to predict the yield of char, liquid and gas products from MSW pyrolysis. The results showed that the coefficients of determination for predicting the yields of char, liquid and gas from the MSW pyrolysis are 0.841, 0.84, and 0.85, respectively; these values demonstrate an accuracy comparable to that achieved for product prediction from single biomass, indicating a successful model performance. The results also show that ash content and temperature are the most important input factors influencing the outputs, namely, yields of char, liquid and gas. The results of this study can help to achieve a more efficient design of the pyrolysis system and improve the recovery of the desired pyrolysis products., Competing Interests: Declaration of competing interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2024 Elsevier Ltd. All rights reserved.)
- Published
- 2024
- Full Text
- View/download PDF