The sensitivity and timeliness of a variation correlation between structures in close proximity and located in the delta region

Abstract

Typical geotechnical structures in the delta region include slopes, foundation pits, and embankments. An analysis of their interrelationships is essential for the overall safety and security of the delta region. At present, these structures are mainly studied by the manual analysis of single monitoring points and by inspection patrols performed by researchers in the field. However, this method is inefficient, costly, and risky to undertake. Furthermore, it is mainly through a time series that the respective states of a single-structure single monitoring point are successfully identified. This makes it difficult to present the correlation between structures and monitoring points, and makes it impossible to truly assess their overall safety patterns. To this end, this study proposes an intelligent analysis method based on feedforward neural networks to associate multiple monitoring points for multiple structures. It uses the correlation between the monitoring values of multiple geotechnical structures and multiple monitoring points for modeling. Then, key coefficients, the gray correlation theory, and degree of importance are introduced to assess the correlation and sensitivity factors of the changes occurring between different structures. The response time and duration between them were also analyzed. Finally, it was applied to a close, typical geotechnical structure system in South China, and the example application was good (R2: 0.91 ∼ 0.95, GR: 0.805 ∼ 0.981). The combined response time and duration of the M index of the foundation pit was optimal for the embankment S (Tr = 0, Tc = 3 ∼ 7) and the JM (Tr = 0, Tc = 4) index of the slope. The multi-correlation model established in this study provides new guidance for accurately revealing the latent hazard potential of typical geotechnical structural systems in the delta region in close proximity. It can also be used as a relevant implication for correlating changes between geotechnical structures in close proximity in other areas.