Landslides are destructive natural disasters that cause catastrophic damage and loss of life worldwide.Accurately predicting landslide displacement enables effective early warning and risk management.However,the limit...Landslides are destructive natural disasters that cause catastrophic damage and loss of life worldwide.Accurately predicting landslide displacement enables effective early warning and risk management.However,the limited availability of on-site measurement data has been a substantial obstacle in developing data-driven models,such as state-of-the-art machine learning(ML)models.To address these challenges,this study proposes a data augmentation framework that uses generative adversarial networks(GANs),a recent advance in generative artificial intelligence(AI),to improve the accuracy of landslide displacement prediction.The framework provides effective data augmentation to enhance limited datasets.A recurrent GAN model,RGAN-LS,is proposed,specifically designed to generate realistic synthetic multivariate time series that mimics the characteristics of real landslide on-site measurement data.A customized moment-matching loss is incorporated in addition to the adversarial loss in GAN during the training of RGAN-LS to capture the temporal dynamics and correlations in real time series data.Then,the synthetic data generated by RGAN-LS is used to enhance the training of long short-term memory(LSTM)networks and particle swarm optimization-support vector machine(PSO-SVM)models for landslide displacement prediction tasks.Results on two landslides in the Three Gorges Reservoir(TGR)region show a significant improvement in LSTM model prediction performance when trained on augmented data.For instance,in the case of the Baishuihe landslide,the average root mean square error(RMSE)increases by 16.11%,and the mean absolute error(MAE)by 17.59%.More importantly,the model’s responsiveness during mutational stages is enhanced for early warning purposes.However,the results have shown that the static PSO-SVM model only sees marginal gains compared to recurrent models such as LSTM.Further analysis indicates that an optimal synthetic-to-real data ratio(50%on the illustration cases)maximizes the improvements.This also demonstrates the robustness and effectiveness of supplementing training data for dynamic models to obtain better results.By using the powerful generative AI approach,RGAN-LS can generate high-fidelity synthetic landslide data.This is critical for improving the performance of advanced ML models in predicting landslide displacement,particularly when there are limited training data.Additionally,this approach has the potential to expand the use of generative AI in geohazard risk management and other research areas.展开更多
Soft computing techniques are becoming even more popular and particularly amenable to model the complex behaviors of most geotechnical engineering systems since they have demonstrated superior predictive capacity,comp...Soft computing techniques are becoming even more popular and particularly amenable to model the complex behaviors of most geotechnical engineering systems since they have demonstrated superior predictive capacity,compared to the traditional methods.This paper presents an overview of some soft computing techniques as well as their applications in underground excavations.A case study is adopted to compare the predictive performances of soft computing techniques including eXtreme Gradient Boosting(XGBoost),Multivariate Adaptive Regression Splines(MARS),Artificial Neural Networks(ANN),and Support Vector Machine(SVM) in estimating the maximum lateral wall deflection induced by braced excavation.This study also discusses the merits and the limitations of some soft computing techniques,compared with the conventional approaches available.展开更多
Geotechnical engineering deals with materials(e.g.soil and rock)that,by their very nature,exhibit varied and uncertain behavior due to the imprecise physical processes associated with their formation(Mitchell and Soga...Geotechnical engineering deals with materials(e.g.soil and rock)that,by their very nature,exhibit varied and uncertain behavior due to the imprecise physical processes associated with their formation(Mitchell and Soga,2005).Modeling the behavior of such materials in geotechnical engineering applications is complex and sometimes beyond the ability of most traditional forms of physically-based engineering methods.In recent years,the application of artificial intelligence(AI)in a wide range of geotechnical engineering has grown rapidly(Nawari et al.,1999;Miranda,2007;Javadi and Rezania,2009;Shahin,2013,2016;Chen et al.,2018;Yin et al.,2018;Jin et al.,2019a,2019b,2019c;Zhang P et al.,2020a).展开更多
Shallow landslide initiation typically results from an interplay of dynamic triggering and preparatory conditions along with static predisposition factors.While data-driven methods for assessing landslide susceptibili...Shallow landslide initiation typically results from an interplay of dynamic triggering and preparatory conditions along with static predisposition factors.While data-driven methods for assessing landslide susceptibility or for establishing rainfall-triggering thresholds are prevalent,integrating spatiotemporal information for dynamic large-area landslide prediction remains a challenge.The main aim of this research is to generate a dynamic spatial landslide initiation model that operates at a daily scale and explicitly counteracts potential errors in the available landslide data.Unlike previous studies focusing on space–time landslide modelling,it places a strong emphasis on reducing the propagation of landslide data errors into the modelling results,while ensuring interpretable outcomes.It introduces also other noteworthy innovations,such as visualizing the final predictions as dynamic spatial thresholds linked to true positive rates and false alarm rates and by using animations for highlighting its application potential for hindcasting and scenario-building.The initial step involves the creation of a spatio-temporally representative sample of landslide presence and absence observations for the study area of South Tyrol,Italy(7400 km2)within well-investigated terrain.Model setup entails integrating landslide controls that operate on various temporal scales through a binomial Generalized Additive Mixed Model.Model relationships are then interpreted based on variable importance and partial effect plots,while predictive performance is evaluated through various crossvalidation techniques.Optimal and user-defined probability cutpoints are used to establish quantitative thresholds that reflect both,the true positive rate(correctly predicted landslides)and the false positive rate(precipitation periods misclassified as landslide-inducing conditions).The resulting dynamic maps directly visualize landslide threshold exceedance.The model demonstrates high predictive performance while revealing geomorphologically plausible prediction patterns largely consistent with current process knowledge.Notably,the model also shows that generally drier hillslopes exhibit a greater sensitivity to certain precipitation events than regions adapted to wetter conditions.The practical applicability of the approach is demonstrated in a hindcasting and scenario-building context.In the currently evolving field of space–time landslide modelling,we recommend focusing on data error handling,model interpretability,and geomorphic plausibility,rather than allocating excessive resources to algorithm and case study comparisons.展开更多
基金supported by the Natural Science Foundation of Jiangsu Province(Grant No.BK20220421)the State Key Program of the National Natural Science Foundation of China(Grant No.42230702)the National Natural Science Foundation of China(Grant No.82302352).
文摘Landslides are destructive natural disasters that cause catastrophic damage and loss of life worldwide.Accurately predicting landslide displacement enables effective early warning and risk management.However,the limited availability of on-site measurement data has been a substantial obstacle in developing data-driven models,such as state-of-the-art machine learning(ML)models.To address these challenges,this study proposes a data augmentation framework that uses generative adversarial networks(GANs),a recent advance in generative artificial intelligence(AI),to improve the accuracy of landslide displacement prediction.The framework provides effective data augmentation to enhance limited datasets.A recurrent GAN model,RGAN-LS,is proposed,specifically designed to generate realistic synthetic multivariate time series that mimics the characteristics of real landslide on-site measurement data.A customized moment-matching loss is incorporated in addition to the adversarial loss in GAN during the training of RGAN-LS to capture the temporal dynamics and correlations in real time series data.Then,the synthetic data generated by RGAN-LS is used to enhance the training of long short-term memory(LSTM)networks and particle swarm optimization-support vector machine(PSO-SVM)models for landslide displacement prediction tasks.Results on two landslides in the Three Gorges Reservoir(TGR)region show a significant improvement in LSTM model prediction performance when trained on augmented data.For instance,in the case of the Baishuihe landslide,the average root mean square error(RMSE)increases by 16.11%,and the mean absolute error(MAE)by 17.59%.More importantly,the model’s responsiveness during mutational stages is enhanced for early warning purposes.However,the results have shown that the static PSO-SVM model only sees marginal gains compared to recurrent models such as LSTM.Further analysis indicates that an optimal synthetic-to-real data ratio(50%on the illustration cases)maximizes the improvements.This also demonstrates the robustness and effectiveness of supplementing training data for dynamic models to obtain better results.By using the powerful generative AI approach,RGAN-LS can generate high-fidelity synthetic landslide data.This is critical for improving the performance of advanced ML models in predicting landslide displacement,particularly when there are limited training data.Additionally,this approach has the potential to expand the use of generative AI in geohazard risk management and other research areas.
基金supported by High-end Foreign Expert Introduction program (No.G20190022002)Chongqing Construction Science and Technology Plan Project (2019-0045)
文摘Soft computing techniques are becoming even more popular and particularly amenable to model the complex behaviors of most geotechnical engineering systems since they have demonstrated superior predictive capacity,compared to the traditional methods.This paper presents an overview of some soft computing techniques as well as their applications in underground excavations.A case study is adopted to compare the predictive performances of soft computing techniques including eXtreme Gradient Boosting(XGBoost),Multivariate Adaptive Regression Splines(MARS),Artificial Neural Networks(ANN),and Support Vector Machine(SVM) in estimating the maximum lateral wall deflection induced by braced excavation.This study also discusses the merits and the limitations of some soft computing techniques,compared with the conventional approaches available.
文摘Geotechnical engineering deals with materials(e.g.soil and rock)that,by their very nature,exhibit varied and uncertain behavior due to the imprecise physical processes associated with their formation(Mitchell and Soga,2005).Modeling the behavior of such materials in geotechnical engineering applications is complex and sometimes beyond the ability of most traditional forms of physically-based engineering methods.In recent years,the application of artificial intelligence(AI)in a wide range of geotechnical engineering has grown rapidly(Nawari et al.,1999;Miranda,2007;Javadi and Rezania,2009;Shahin,2013,2016;Chen et al.,2018;Yin et al.,2018;Jin et al.,2019a,2019b,2019c;Zhang P et al.,2020a).
基金The research leading to these results is related to the PROSLIDE project that received funding from the research program Research Südtirol/Alto Adige 2019 of the Autonomous Province of Bozen/Bolzano-Südtirol/Alto Adige.
文摘Shallow landslide initiation typically results from an interplay of dynamic triggering and preparatory conditions along with static predisposition factors.While data-driven methods for assessing landslide susceptibility or for establishing rainfall-triggering thresholds are prevalent,integrating spatiotemporal information for dynamic large-area landslide prediction remains a challenge.The main aim of this research is to generate a dynamic spatial landslide initiation model that operates at a daily scale and explicitly counteracts potential errors in the available landslide data.Unlike previous studies focusing on space–time landslide modelling,it places a strong emphasis on reducing the propagation of landslide data errors into the modelling results,while ensuring interpretable outcomes.It introduces also other noteworthy innovations,such as visualizing the final predictions as dynamic spatial thresholds linked to true positive rates and false alarm rates and by using animations for highlighting its application potential for hindcasting and scenario-building.The initial step involves the creation of a spatio-temporally representative sample of landslide presence and absence observations for the study area of South Tyrol,Italy(7400 km2)within well-investigated terrain.Model setup entails integrating landslide controls that operate on various temporal scales through a binomial Generalized Additive Mixed Model.Model relationships are then interpreted based on variable importance and partial effect plots,while predictive performance is evaluated through various crossvalidation techniques.Optimal and user-defined probability cutpoints are used to establish quantitative thresholds that reflect both,the true positive rate(correctly predicted landslides)and the false positive rate(precipitation periods misclassified as landslide-inducing conditions).The resulting dynamic maps directly visualize landslide threshold exceedance.The model demonstrates high predictive performance while revealing geomorphologically plausible prediction patterns largely consistent with current process knowledge.Notably,the model also shows that generally drier hillslopes exhibit a greater sensitivity to certain precipitation events than regions adapted to wetter conditions.The practical applicability of the approach is demonstrated in a hindcasting and scenario-building context.In the currently evolving field of space–time landslide modelling,we recommend focusing on data error handling,model interpretability,and geomorphic plausibility,rather than allocating excessive resources to algorithm and case study comparisons.
