Article: New algorithm for numerical simulation of beach evolution under extreme weather and neural network optimisation prediction model Journal: International Journal of Environment and Pollution (IJEP) 2025 Vol.75 No.4 pp.280 - 299 Abstract: This study proposes a novel beach evolution prediction algorithm integrating convolutional neural networks and numerical simulation to enhance accuracy under extreme weather. An improved deep-water flow model, based on the Navier-Stokes and sand-water mixing equations, captures hydrodynamic changes influenced by wind, waves, tides, and currents. Meteorological and oceanic data are preprocessed using local weighted regression and interpolation methods to ensure quality. A neural network model dynamically predicts beach evolution, with k-fold cross-validation ensuring stability across extreme weather scenarios. Results show high accuracy, with mean square error (MSE), root mean square error (RMSE), and mean absolute error (MAE) all below 0.4 and prediction errors under 12%. Inderscience Publishers - linking academia, business and industry through research

Title: New algorithm for numerical simulation of beach evolution under extreme weather and neural network optimisation prediction model

Authors: Songzhe Li; Hongqian Zhang

Addresses: Key Laboratory of Engineering Sediment, Tianjin Research Institute for Water Transport Engineering, Ministry of Transport, Tianjin, 300456, China ' Key Laboratory of Engineering Sediment, Tianjin Research Institute for Water Transport Engineering, Ministry of Transport, Tianjin, 300456, China

Abstract: This study proposes a novel beach evolution prediction algorithm integrating convolutional neural networks and numerical simulation to enhance accuracy under extreme weather. An improved deep-water flow model, based on the Navier-Stokes and sand-water mixing equations, captures hydrodynamic changes influenced by wind, waves, tides, and currents. Meteorological and oceanic data are preprocessed using local weighted regression and interpolation methods to ensure quality. A neural network model dynamically predicts beach evolution, with k-fold cross-validation ensuring stability across extreme weather scenarios. Results show high accuracy, with mean square error (MSE), root mean square error (RMSE), and mean absolute error (MAE) all below 0.4 and prediction errors under 12%.

Keywords: extreme weather; beach evolution; numerical simulation; neural network; prediction analysis.

DOI: 10.1504/IJEP.2025.150933

International Journal of Environment and Pollution, 2025 Vol.75 No.4, pp.280 - 299

Received: 11 Feb 2025
Accepted: 05 Jun 2025
Published online: 05 Jan 2026 *

Title: New algorithm for numerical simulation of beach evolution under extreme weather and neural network optimisation prediction model

Keep up-to-date