Sustainable Infrastructure Systems For Green Buildings: Lifecycle Optimization And Performance Modeling Using Machine Learning

Abstract

Resource-efficient designs, lifecycle assessment, and intelligent control mechanisms characterize sustainable infrastructures that minimize the environmental impacts while maintaining functionality. Nevertheless, current methodologies are unable to optimize the lifecycle process, predict future events, and incorporate sustainability objectives. This research focuses on developing an innovative machine learning (ML) system for lifecycle optimization and performance modeling of green buildings. The research aims to promote sustainability, energy efficiency, and occupant comfort in green buildings. The Green Build Lifecycle Performance data consists of 9,400 entries having 40 attributes is adopted for this research. Energy consumption, environmental characteristics, occupancy behaviors, and other sustainability variables are recorded in the dataset. Data pre-processing is performed through Min-Max normalization, and the most important features are extracted using Principal Component Analysis (PCA). The dataset is employed to train predictive models and improve decision-making throughout the building lifecycle. An Efficient Ant Colony-tuned Dynamic Random Forest Classifier (EAC-DRFC) is proposed for predictive models. EAC helps select features and tune hyperparameters for the EAC-DRFC model, whereas DRFC helps learn adaptively based on the dynamic nature of the data. The proposed model using Python is impressive in terms of its accuracy of 0.97, precision of 0.96, recall of 0.98, and F1-score of 0.97. The research offers intelligent, adaptive, and lifecycle-oriented management of sustainable green building infrastructure systems.