Nowadays, waste to energy (WTE) transformation solutions play a vital role in waste disposal. Accurate WTE resource planning can be made using high-performance waste amount prediction models. Thus, a significant gain can be obtained both in economic and environmental terms. In this paper, we proposed different machine learning models to predict the amount of municipal solid waste (MSW) to be used for smart energy management systems. To point this problem, we study a new WTE Framework and use the real-world data set obtained from MSW stations on the European side of Istanbul, Turkey. The basis of our motivation for choosing Istanbul is based on the ‘Waste Incineration and Power Generation Plant,¹’ which was built in Eyupsultan, Istanbul in 2017 and is planned to be operational in 2021. This plant will be Europe’s largest domestic waste incinerator with a capacity of 3000 tons/day. For the proposed WTE framework, we first build an ensemble model, Gradient Boosting (GB), to predict the amount of MSW using daily data related to other variables such as seasonality and socio-economic status. Then we use the calorific index value to predict generated energy from solid waste, categorized in 14 different waste types.

如今，废物转化为能源（WTE）的转换解决方案在废物处理中起着至关重要的作用。可以使用高性能浪费量预测模型来进行准确的WTE资源计划。因此，在经济和环境方面都可以获得显着的收益。在本文中，我们提出了不同的机器学习模型来预测将用于智能能源管理系统的城市固体废物（MSW）的数量。为了解决这个问题，我们研究了一个新的WTE框架，并使用从土耳其伊斯坦布尔欧洲一侧的MSW站获得的真实数据集。我们选择伊斯坦布尔的动机基于“废物焚化和发电厂¹该厂于2017年在伊斯坦布尔的Eyupsultan建成，计划于2021年投入运营。该厂将成为欧洲最大的生活垃圾焚烧炉，日处理能力为3000吨。对于拟议的WTE框架，我们首先建立一个集成模型Gradient Boosting（GB），以使用与其他变量（例如季节性和社会经济状况）相关的每日数据来预测MSW的数量。然后，我们使用发热量指数值来预测固体废物产生的能量，将其分类为14种不同的废物类型。