ISSN 1000-0526
CN 11-2282/P
CN 11-2282/P
Research on Pavement Temperature Forecasting Method of Tianjin Expressway based on Long Short-Term Memory Network
- Article
- Figures
- Metrics
- Preview PDF
- Reference
- Related
- Cited by
- Materials
Abstract:
Based on the 10-minutes observational data from 2018 to 2023 at the traffic meteorological stations along the Tianjin expressway, a long short-term memory network model was used to forecast the pavement temperature in the next 3 hours, and the forecasting effect and spatial migration applicability of the model in the case of extreme pavement temperature were evaluated and analyzed. The results show that: the model forecasting effect is the best when the observational data of the past 3 hours length is used as the input; the model can forecast the pavement temperature accurately, and the forecasting ability decreases with the extension of the forecasting time limit, with the mean absolute deviation of 0.38-2.21℃, the root-mean-square error of 0.63-3.31℃, and the accuracy rate of 76.15%-99.48%; models can accurately forecast the timing and extremes of extreme low temperatures occurring on pavements, forecast accuracy control of about 90% for the 1-hour forecast time frame; for extreme pavement temperatures, the model"s 1-hour forecast also simulates the trend and time to extremes in pavement temperatures; the model has a certain spatial migration ability, and the average accuracy of the model forecast is more than 70% after the migration, with an mean absolute deviation of no more than 3℃; within a certain range, compared with the migration distance, the forecasting ability of the model"s home station has a greater impact on the forecasting performance of the model after migration.
Keywords:
Project Supported:
Clc Number:
Mobile website
