Telescopic boom cranes make it difficult to monitor the full-domain stress in real operations due to the complex structure and randomly variable loads and environmental conditions. A structural performance prediction method based on digital twin is proposed, which combines real-time monitoring data with finite element simulation data and incorporates telescopic boom expertise with feature derivation, utilizes the XGBoost algorithm to establish a real-time stress prediction agent model, realizes the structural stress prediction of telescopic boom cranes with the boom and rotary table as the main key components of the whole life cycle, and builds a digital twin system of telescopic boom crane structural performance using Unity 3D to realize real-time stress control and real-time stress prediction, and builds a digital twin system of telescopic boom crane structural performance. Unity 3D is used to build a digital twin system of telescopic boom crane structural performance, realizing mutual control of reality and real-time stress cloud display to ensure its safety performance.