Title: Forecasting municipal solid waste generation in Tianjin based on long and short term memory neural network model

Authors: Xiaoqin Liang; Bingchun Liu; Xin Qi; Jie Ji; Xiaogang Yu

Addresses: School of Management, Tianjin University of Technology, Tianjin, 300384, China ' School of Management, Tianjin University of Technology, Tianjin, 300384, China ' School of Management, Tianjin University of Technology, Tianjin, 300384, China ' School of Management, Tianjin University of Technology, Tianjin, 300384, China ' Tianjin 712 Mobile Communication Co., Ltd., Tianjin, 300140, China

Abstract: Accurately forecasting urban waste collection volume is of great significance for formulating waste treatment policies. This study focuses on Tianjin and constructs a prediction model based on the long short-term memory neural network (LSTM) model, using seven factors as input indicators affecting the generation of urban domestic waste, including natural gas, artificial gas, liquefied petroleum gas, water supply, population, gross domestic product (GDP), and total social retail sales. The experimental results demonstrate that the LSTM model established in this study is accurate in predicting waste volume, with a mean absolute percentage error (MAPE) of 9.27. Furthermore, by predicting the future trend of waste generation under different scenarios, it is found that municipal waste generation in Tianjin will be between 2.3918 MT and 5.1340 MT during 2019-2023. Finally, this study proposes relevant suggestions to deal with the increase in the amount of municipal solid waste.

Keywords: deep learning; LSTM; long short-term memory; neural networks; forecasting; municipal solid waste; sustainable development; pollution; environment; health; landfills.

DOI: 10.1504/IJEP.2021.132008

International Journal of Environment and Pollution, 2021 Vol.70 No.3/4, pp.223 - 240

Received: 30 Dec 2021
Received in revised form: 16 Nov 2022
Accepted: 18 Nov 2022
Published online: 06 Jul 2023 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article