Search

Based on the machine learning algorithms, prediction models for landslide displacement with step-like behavior in the reservoir area were established for landslides prevention and reduction; these models could predict a given test set very well. However, due to the length and the sequence of the training set in prediction models, the predictive ability of these prediction models could not be evaluated accurately if just validated with a given test set. To solve the problem, a hybrid reliability model was proposed. Complimentary ensemble empirical mode decomposition (CEEMD) algorithm was used to decompose the accumulated displacement into the trend displacement and the periodic displacement firstly. The Gauss function was used to predict the trend displacement, and the random forest (RF) and back propagation neural network (BPNN) algorithms were employed to predict the periodic displacement. Furthermore, a novel performance function for the reliability analysis of the displacement prediction model was derived to address failure probabilities in different cases. The Baijiabao landslide in the Three Gorges Reservoir Area was taken as an example for reliability analysis of the prediction model for landslide displacement with step-like behavior, and the predictive ability of the CEEMD-RF model and the CEEMD-BPNN model were compared. The results indicated that the CEEMD-RF model and CEEMD-BPNN model both can accurately predict the accumulated displacement of the given test set; the predictive value obtained with the CEEMD-RF model or the CEEMD-BPNN model showed uncertainties of the prediction model for landslide displacement and the predictive ability of the CEEMD-RF model was more reliable than the CEEMD-BPNN model under different cases. The failure probability proposed in the paper could evaluate the predictive ability of the model more accurately and comprehensively compared with the existing assessment indices. [ABSTRACT FROM AUTHOR]

Understandings of reinforcement mechanisms of landslide-stabilizing pile system are important for long-term safety of reservoir landslides installed piles. The paper proposes a framework to study deformation response and identify triggering factors for landslide–pile system by using geographic detector technology and uncertainty of monitoring data. Majiagou landslide, a representative reservoir landslide installed stabilizing and test piles, is selected as the case study. Firstly, monitoring data of monthly rainfall, variations of reservoir water level, deformation of landslide surface and piles' head were preprocessed. The random deformation data were generated considering uncertainty of deformation monitoring data. Meanwhile, the deformation response of landslide–pile system is analyzed through studying the monitoring deformation data and random deformation data using the clustering algorithm. Finally, geographic detector technique was used to identify main triggering factors of landslide surface deformation and explore interaction types of any two factors. The uncertainty of monitoring data was used to identify the most important triggering factor. Influences of different error and uncertainty of monitoring data on the influence degrees of each factor were further discussed. The comparison with improved Apriori algorithm shows that the presented framework can intuitively measure influence degrees of each factor and analyze interaction types of any two factors. The main conclusions by studying Majiagou landslide indicate that the anti-slide performance and action range of piles gradually decrease with the increase of hydraulic cycle; the deterioration of geomaterials' properties are the most important triggering factor leading to the deformation of landslide–pile system and the degradation of piles' performance, supported by most random deformation groups. [ABSTRACT FROM AUTHOR]

Zigui Basin is a major landslide-prone region in the Three Gorges Reservoir region of China, and the stabilizing pile is an effective and widely employed countermeasure to reinforce landslides in this region. However, stabilizing piles are mostly designed using deterministic and stability-oriented methods, which generally ignore the system performance and cost-effectiveness. Using the Majiagou landslide reinforced with stabilizing piles as a case study, a probabilistic multi-objective optimization framework for the design of stabilizing piles is proposed and illustrated. Specifically, performance objectives related to failure probability, system robustness and life-cycle cost of the landslide-stabilizing pile system with feasible designs are evaluated, then the best compromised design is obtained by means of Pareto optimality. Expert knowledge and professional judgment are required to set necessary restrictions and finally determine the optimal design. The results show that there is a better design of stabilizing piles than the existing one, with which acceptable reinforcement effectiveness, compromised life-cycle cost and robust system performance can be realized. The optimal design will also vary with the concerned performance objectives and knowledge-based judgment. Further relationships and interpretations between design parameters and system responses are discussed through parametric analyses. [ABSTRACT FROM AUTHOR]