{"title":"利用综合气候和地温数据加强人工冻土层表预测:青藏公路案例研究","authors":"Yu-Zhi Zhang , Shao-Jie Liang , Jian-Bing Chen , Meng Wang , Ming-Tao Jia , Ya-Ting Jiang","doi":"10.1016/j.coldregions.2024.104341","DOIUrl":null,"url":null,"abstract":"<div><div>The complexities of permafrost changes, driven by climate warming and engineering activities, coupled with challenges in data acquisition, make it crucial and challenging to accurately predict the artificial permafrost table, particularly for subgrades in high-temperature unstable permafrost regions. To address this, this study developed a hybrid machine learning model (RF-LSTM-XGBoost) for permafrost table prediction. By analyzing climate change and ground temperature data from various positions and depths along the subgrade in the Tuotuo River section of the Qinghai-Xizang Highway, the Spearman correlation coefficient method was initially used to determine the important influencing factors. Random Forest (RF), Long Short-Term Memory Neural Network (LSTM), and Extreme Gradient Boosting (XGBoost) models were used to predict the artificial permafrost table, and grid search and cross-validation methods were employed to optimize the hyperparameters of each model. A linear weighted combination method based on the minimum cumulative absolute error was utilized to merge the models, and its performance was compared with the individual RF, LSTM, and XGBoost models. Subsequently, the feature importance of the variables in the machine learning model was analyzed. The results indicated a strong correlation between artificial permafrost table changes and factors such as daily average atmospheric temperature, subgrade surface ground temperature, and subgrade surface ground heat flux during the freezing-thawing cycle. The combined model highlighted daily atmospheric temperature as the most influential predictor, followed by ground heat flux, with the surface ground temperature being less impactful. The combined model demonstrated improved predictive accuracy, with MSE, MAPE, RMSE, MAE, and R<sup>2</sup> values of 0.003, 0.052, 0.0085, 0.029, and 0.989, respectively, surpassing those of individual models. This model offers a rapid, accurate, and reliable approach for permafrost table prediction, advancing subgrade stability research in challenging permafrost environments.</div></div>","PeriodicalId":10522,"journal":{"name":"Cold Regions Science and Technology","volume":"229 ","pages":"Article 104341"},"PeriodicalIF":3.8000,"publicationDate":"2024-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Enhancing artificial permafrost table predictions using integrated climate and ground temperature data: A case study from the Qinghai-Xizang highway\",\"authors\":\"Yu-Zhi Zhang , Shao-Jie Liang , Jian-Bing Chen , Meng Wang , Ming-Tao Jia , Ya-Ting Jiang\",\"doi\":\"10.1016/j.coldregions.2024.104341\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<div><div>The complexities of permafrost changes, driven by climate warming and engineering activities, coupled with challenges in data acquisition, make it crucial and challenging to accurately predict the artificial permafrost table, particularly for subgrades in high-temperature unstable permafrost regions. To address this, this study developed a hybrid machine learning model (RF-LSTM-XGBoost) for permafrost table prediction. By analyzing climate change and ground temperature data from various positions and depths along the subgrade in the Tuotuo River section of the Qinghai-Xizang Highway, the Spearman correlation coefficient method was initially used to determine the important influencing factors. Random Forest (RF), Long Short-Term Memory Neural Network (LSTM), and Extreme Gradient Boosting (XGBoost) models were used to predict the artificial permafrost table, and grid search and cross-validation methods were employed to optimize the hyperparameters of each model. A linear weighted combination method based on the minimum cumulative absolute error was utilized to merge the models, and its performance was compared with the individual RF, LSTM, and XGBoost models. Subsequently, the feature importance of the variables in the machine learning model was analyzed. The results indicated a strong correlation between artificial permafrost table changes and factors such as daily average atmospheric temperature, subgrade surface ground temperature, and subgrade surface ground heat flux during the freezing-thawing cycle. The combined model highlighted daily atmospheric temperature as the most influential predictor, followed by ground heat flux, with the surface ground temperature being less impactful. The combined model demonstrated improved predictive accuracy, with MSE, MAPE, RMSE, MAE, and R<sup>2</sup> values of 0.003, 0.052, 0.0085, 0.029, and 0.989, respectively, surpassing those of individual models. This model offers a rapid, accurate, and reliable approach for permafrost table prediction, advancing subgrade stability research in challenging permafrost environments.</div></div>\",\"PeriodicalId\":10522,\"journal\":{\"name\":\"Cold Regions Science and Technology\",\"volume\":\"229 \",\"pages\":\"Article 104341\"},\"PeriodicalIF\":3.8000,\"publicationDate\":\"2024-10-18\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Cold Regions Science and Technology\",\"FirstCategoryId\":\"5\",\"ListUrlMain\":\"https://www.sciencedirect.com/science/article/pii/S0165232X24002222\",\"RegionNum\":2,\"RegionCategory\":\"工程技术\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"ENGINEERING, CIVIL\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Cold Regions Science and Technology","FirstCategoryId":"5","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0165232X24002222","RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ENGINEERING, CIVIL","Score":null,"Total":0}
Enhancing artificial permafrost table predictions using integrated climate and ground temperature data: A case study from the Qinghai-Xizang highway
The complexities of permafrost changes, driven by climate warming and engineering activities, coupled with challenges in data acquisition, make it crucial and challenging to accurately predict the artificial permafrost table, particularly for subgrades in high-temperature unstable permafrost regions. To address this, this study developed a hybrid machine learning model (RF-LSTM-XGBoost) for permafrost table prediction. By analyzing climate change and ground temperature data from various positions and depths along the subgrade in the Tuotuo River section of the Qinghai-Xizang Highway, the Spearman correlation coefficient method was initially used to determine the important influencing factors. Random Forest (RF), Long Short-Term Memory Neural Network (LSTM), and Extreme Gradient Boosting (XGBoost) models were used to predict the artificial permafrost table, and grid search and cross-validation methods were employed to optimize the hyperparameters of each model. A linear weighted combination method based on the minimum cumulative absolute error was utilized to merge the models, and its performance was compared with the individual RF, LSTM, and XGBoost models. Subsequently, the feature importance of the variables in the machine learning model was analyzed. The results indicated a strong correlation between artificial permafrost table changes and factors such as daily average atmospheric temperature, subgrade surface ground temperature, and subgrade surface ground heat flux during the freezing-thawing cycle. The combined model highlighted daily atmospheric temperature as the most influential predictor, followed by ground heat flux, with the surface ground temperature being less impactful. The combined model demonstrated improved predictive accuracy, with MSE, MAPE, RMSE, MAE, and R2 values of 0.003, 0.052, 0.0085, 0.029, and 0.989, respectively, surpassing those of individual models. This model offers a rapid, accurate, and reliable approach for permafrost table prediction, advancing subgrade stability research in challenging permafrost environments.
期刊介绍:
Cold Regions Science and Technology is an international journal dealing with the science and technical problems of cold environments in both the polar regions and more temperate locations. It includes fundamental aspects of cryospheric sciences which have applications for cold regions problems as well as engineering topics which relate to the cryosphere.
Emphasis is given to applied science with broad coverage of the physical and mechanical aspects of ice (including glaciers and sea ice), snow and snow avalanches, ice-water systems, ice-bonded soils and permafrost.
Relevant aspects of Earth science, materials science, offshore and river ice engineering are also of primary interest. These include icing of ships and structures as well as trafficability in cold environments. Technological advances for cold regions in research, development, and engineering practice are relevant to the journal. Theoretical papers must include a detailed discussion of the potential application of the theory to address cold regions problems. The journal serves a wide range of specialists, providing a medium for interdisciplinary communication and a convenient source of reference.