Abstract: To accurately and reasonably evaluate the service status of urban metro shield tunnels, this paper puts for? ward a dynamic evaluation method for tunnel performance. The method includes two parts: (1) knowledge graph of shield tunnels; and (2) dynamic evaluation model of tunnel performance. The Knowledge Graph (KG) is used to integrate data related to shield tunnels and provide data support for the evaluation of shield tunnel performance. In terms of the performance evaluation model, the cumulative difference approach (CDA) is used to divide the tunnel into different segments of unequal lengths based on the lateral convergence deformation of the shield tunnel, taking into account various factors such as data scarcity, passages, and the influence extent of failures, and the segmentation results are adjusted through dynamic time warping (DTW); Based on the segmentation results, the static evaluation results of the shield tunnel′ s performance during multiple consecutive time intervals are used to calculate the rate and trend of changes in the performance. During the application of the method in Shanghai Metro Line 1, the model accurately identified homogeneous sections and tunnel cross-sections with different deterioration rates, and the KG also provided an effective approach for the application of the model.