Abstract

Abstract

Abstract

Accurately obtaining the mechanical properties of embankment soil is the core component for achieving precise calculation and mastering the operational status of the embankment. To achieve precise identification of multiple key material parameters of the embankment soil, this paper proposes an inversion method that integrates monitoring displacement information with intelligent algorithms. This approach comprehensively utilizes monitoring data from multiple monitoring points and different times to collaboratively invert multiple key parameters of the embankment soil. The study establishes a finite element model of the embankment based on the Cvisc (Burgers-Mohr) viscoelastic-plastic constitutive theory, which describes the creep behavior of soil. Orthogonal experimental design is employed to generate multiple parameter combinations of the soil, and numerical simulations are conducted to obtain displacement increments at various monitoring points over time, thereby forming a training sample set. On this basis, using the displacement increment sequences from multiple monitoring points at different times as input, a Back Propagation (BP) neural network improved by the Whale Optimization Algorithm (WOA) is trained to establish a nonlinear mapping model from displacement sequences to material parameters. Subsequently, the material parameters of the soil are inverted based on the actual displacement sequences from multiple monitoring points at different times. This method is applied to the Tongma Embankment for parameter inversion. Verification shows that the displacements calculated using the inverted parameters are in good agreement with the monitoring values, with an average relative error of approximately 2.8%. The results indicate that this method can effectively invert multiple parameters of the soil, thereby providing a reliable basis for precisely understanding the mechanical properties of the soil and the state of the embankment.OPEN ACCESS Received: 17/10/2025 Accepted: 28/11/2025 Published: 21/09/2026

Abstract

Electrocardiogram (ECG) signal processing plays a critical role in the early detection and diagnosis of cardiovascular diseases; however, reliable automated interpretation remains challenging due to noise contamination, inter-patient variability, and nonstationary signal characteristics. This review provides a critical and structured analysis of numerical methods and computational algorithms used in ECG signal processing and cardiac abnormality detection. Unlike conventional descriptive surveys, this work emphasizes the comparative evaluation of preprocessing techniques, feature extraction strategies, and machine learning and deep learning models under varying conditions such as noise intensity, dataset size, and realtime deployment constraints. The review systematically examines baseline correction, filtering, wavelet and decomposition-based denoising, fiducial point detection, and feature representation methods, highlighting their computational trade-offs, robustness, and clinical applicability. Furthermore, it critically contrasts classical machine learning approaches with modern deep learning and transformer-based models, identifying scenarios in which each paradigm is most effective. Key limitations of widely used datasets, particularly the MIT-BIH Arrhythmia Database, are analyzed in terms of generalization and evaluation bias. Emerging challenges, including domain shift, annotation inconsistency, interpretability, and deployment in wearable systems, are discussed with concrete research directions such as hybrid modeling, explainable AI, and federated learning. The review concludes that high-performance ECG analysis systems require an integrated pipeline combining numerically stable preprocessing with adaptive and interpretable learning frameworks, rather than reliance on standalone classification models.