Author(s)

Xiayun Lin

Corresponding Author

Xiayun Lin

Abstract

The traditional decision-making method of construction project cost depends on manual experience and static data model, which has some problems such as inefficient data utilization, lagging dynamic response and weak risk prediction. This study constructs a dynamic cost decision support system (DSS) based on artificial intelligence, aiming at solving the above problems and promoting the digital transformation of the construction industry to data-driven and intelligent decision-making. The system adopts hierarchical architecture, including data layer, algorithm layer, application layer and interaction layer, and realizes the whole process from multi-source data perception to intelligent decision-making suggestion. The data layer is responsible for the collection, fusion and governance of multimodal data, and is deeply integrated with BIM model. The algorithm layer includes dynamic cost prediction, risk identification and early warning, multi-objective optimization decision-making and other modules, and uses deep learning, transfer learning, reinforcement learning and other technologies to model and optimize. The application layer provides real-time cost kanban, risk early warning center, intelligent scheme optimizer and other functional modules to provide decision support for different users. The interactive layer presents the system results with a visual and interactive interface to ensure the effective transmission of decision suggestions. The case verification shows that the system can significantly improve the accuracy of cost prediction, realize early detection and early warning of risks, generate quantitative and feasible optimization schemes, and effectively assist managers to make scientific decisions. This study provides a new theoretical framework and practical tools for the intersection of AI and engineering management, and has broad application prospects.

Keywords

Artificial Intelligence, Construction Project Cost, Decision Support System