Akademik Veri Yönetim Sistemi
Tezin Türü: Doktora
Tezin Yürütüldüğü Kurum: Yıldız Teknik Üniversitesi, İnşaat Fakültesi, İnşaat Mühendisliği Bölümü, Türkiye
Tezin Onay Tarihi: 2025
Tezin Dili: İngilizce
Öğrenci: Ruba M. A. Awad
Asıl Danışman (Eş Danışmanlı Tezler İçin): Aslı Pelin Gürgün
Açık Arşiv Koleksiyonu: AVESİS Açık Erişim Koleksiyonu
Özet:
Construction and demolition (C&D) projects rely on predicting waste generation hence an accurate predicting model assists the scheduling process by saving time, energy, and cost. This research will develop a hybrid model by coupling pre-processing techniques as Archimedes optimization algorithm (AOA), and gray wolf optimization (GWO) algorithm, and a machine learning algorithm such as random forest (RF), and artificial neural network (ANN) to predict the waste generated in construction projects. The study's concept of a hybrid model will be feature selection. AOA and GWO were employed as a feature selection for the hybrid model which select the best input combinations according to minimum error and maximum coefficient of determination (R2). For the development of the model, data on 200 real-life C&D projects from Gaza Strip were used. Demographic and statistical analysis were used for data analysis. The efficiency of the developed models was evaluated using different performance parameters including the mean absolute error (MAE), the mean square error (MSE), the root mean squared error (RMSE), and the coefficient of determination (R2). The results demonstrated that the AOA-ANN model presented better data modeling than the classical ANN, RF model and hybridized AOA-RF, GWO-ANN and GWO-RF models. The results of testing phase were (e.g., MAE, RMSE, and MSE) (0.0086648, 0.023728, and 0.00056304) and an R2 value of (0.99333) for the AOA-ANN (model 5). This was an attempt to implement the new intelligent techniques in waste management area. Moreover, it can be used to help project managers to control project time, cost and make accurate waste amount predictions.