基于Informer深度学习的桥梁温度场预测方法
By constructing a bridge temperature field prediction method using the Informer deep learning model, the problem of traditional methods being unable to accurately predict bridge temperature fields is solved, achieving efficient and reliable bridge temperature field prediction, and improving bridge safety and maintenance efficiency.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- 中电建路桥集团有限公司
- Filing Date
- 2025-10-13
- Publication Date
- 2026-07-17
AI Technical Summary
Traditional methods are insufficient to effectively and promptly detect internal temperature defects in bridges. Existing bridge temperature field prediction methods are also unable to accurately predict future temperature fields when weather conditions change drastically, affecting bridge safety and maintenance efficiency.
A bridge temperature field prediction method based on Informer deep learning is adopted. By constructing a sample library containing time features, environmental features and temperature, and integrating meteorological data to train the Informer model, the time series prediction of the temperature field of concrete beam bridge is realized. The sequence prediction is performed by using the ProbSparse self-attention mechanism and multi-head ProbSparse attention layer.
It has achieved long-term sustainable time-series prediction of bridge temperature field, significantly improved prediction efficiency, and the predicted values are in high agreement with the actual observed values, providing efficient and reliable temperature effect assessment and safety early warning support.
Smart Images
Figure CN121303200B_ABST