In order to enhance the intelligence level of construction engineering

research on intelligent construction of building based on data fusion and state perception is conducted. This paper analyzes the fusion technology of multi-source heterogeneous data and proposes a hybrid decision fusion framework based on deep learning and evidence theory. For the monitoring of construction equipment status

a multi-sensor collaborative perception and residual network diagnostic model are adopted to achieve real-time evaluation of equipment health status. For the situation awareness of environmental safety

a dynamic risk field model is established

and combined with BIM and IoT data

hazard warnings are realized. In the field of new unmanned digital construction

the paper explores the robot path planning for intelligent masonry and the multi machine collaborative control strategy for unmanned earthwork engineering. Experimental verification and case analysis show that the proposed method can effectively improve the automation and safety of the construction process

providing a feasible technical path for the practical application of intelligent construction.