Intelligent optimization-based calibration of DEM numerical parameters using an improved sparrow search algorithm

Abstract Calibration of meso parameters is the foundation of discrete element method (DEM) simulations. To overcome the limitations of the traditional trial-and-error calibration method, such as its time-consuming process and poor repeatability, this study proposes an efficient DEM meso parameters calibration method based on an improved sparrow search algorithm (ISSA). The algorithm integrates multiple enhancement strategies, including improved circle chaotic mapping for population initialization, adaptive inertia weight, Lévy flight, and Cauchy–Gaussian mutation. The results demonstrate that the proposed calibration method is applicable to multi-scale models and accurately reproduces both laboratory-scale specimen results and field-scale experimental observations. The laboratory-scale DEM simulation achieved parameter calibration with a total error of less than 1% within 4.4 hours, while the field-scale DEM simulation results showed discrepancies of only 0.21% in failure depth and 0.29% in failure range compared with field observations. Furthermore, under the same accuracy conditions, the ISSA exhibits a remarkable speed advantage in meso parameters calibration, outperforming the SSA-DEM, WOA-DEM, DA-DEM, and PSO-DEM algorithms by 83.33%, 128.57%, 226.19%, and 371.42%, respectively. These results confirm that the proposed calibration method achieves automated and efficient parameter identification for DEM under the guidance of intelligent optimization algorithms.