Dual-Dimensional Collaborative Analysis for the Optimization of Slope Monitoring Points

Abstract The frequent occurrence of landslide hazards poses severe threats to infrastructure safety and human life. Traditional monitoring point deployment primarily relies on engineering experience, often leading to monitoring blind spots or resource redundancy. This study proposes a dual-dimensional collaborative framework that integrates macroscopic temporal trend similarity with microscopic chaotic dynamic state similarity to identify homogeneous dynamic response zones (HDRZs) for monitoring layout optimization. The core methodological innovations include (1) an improved rise angle–average frequency method featuring a sliding-window adaptive threshold algorithm and subdivided frequency parameters for low-frequency displacement series analysis, and (2) a collaborative decision rule that couples temporal evolution (quantified via dynamic time warping) with intrinsic dynamic sensitivity (characterized by the maximum Lyapunov exponent). Using FLAC3D (version 7.0) numerical simulations as a digital twin for predictive design, the proposed framework partitions the slope into homogeneous dynamic response zones and reduces monitoring redundancy while maintaining coverage of critical geomechanical regions such as rear-edge fractures and potential shear outlets. The engineering applicability of the proposed framework was further examined through a case study on the G5 Expressway by comparing the algorithm-derived zoning results with an existing expert-designed monitoring layout, where a high degree of spatial consistency was observed between the identified response zones and the empirical monitoring scheme. Crucially, the identified HDRZs encapsulate power-law acceleration characteristics, providing high-fidelity data that support failure forecasting within the framework of the time-reversed Omori law. This research provides a quantitative framework for monitoring layout optimization in complex slope engineering.