Forecasting of municipal solid waste multi-classification by using time-series deep learning depending on the living standard

The type and quantity of municipal solid waste are important factors for determining how these wastes should be handled, managed, and valorised. This paper investigates the effect of different living styles on the type of generated municipal solid waste (MSW). It is also forecasting the amount and type of generated municipal solid waste. Al-basaten, a district at East-Cairo, Egypt was considered a case study due to the diversity of lifestyles. The Al-basaten area has three different zones depending on level styles: poor, social, and privileged zones. Solid waste was collected separately from each zone, sorted (as plastic, glass, paper, carton, and organic waste), gathered the same type of sorted solidwaste from each zone individually, and weighted. The analysis of dis-tinguishing waste is discussed. The forecasting model by using a long short-term memory (LSTM) along with deep learning time series forecasting (DLTSF) network was used for Al-basaten MSW. The forecasting model was trained, validated, and tested on the real-life dataset of the sorted and weighted waste from each zone. The analyzed results provide the average solid waste was 0.42, 0.65, and 0.86 kg/person/day for poor, social, and privileged zones, respectively. The forecasting results indicated that the proposed model could effectively forecast the future series values of the plastic, glass, paper, carton, and organic waste for the poor, social, and privileged zones. The mean RMSE of DLTSF was 0.03371 for forecasting the total MSW. This analysis will help decision-makers maximize solidwaste recycling benefits.

Automated summary

This paper investigates the impact of different living styles on the type and quantity of municipal solid waste (MSW) and proposes a forecasting model using deep learning techniques to predict the future waste generation. The results show that the average solid waste generation varies among different areas with different living styles. The forecasting model effectively predicts the future waste composition for each area. This analysis provides valuable insights for decision-makers to optimize solid waste recycling.