GMHEA: A New Software for the Multi-Variable Optimization of High-Entropy Alloy by the Improved Genetic Algorithm and Effective Medium Theory.

(x = 1~10) and addresses key limitations of existing methods, including low search efficiency, high computational cost, and inadequate sampling of configurational space. GMHEA incorporates adaptive genetic operators (cut-and-splice pairing, soft mutation, strain mutation), structure comparison based on fingerprint, and variable-cell relaxation to ensure robust convergence to the stable structures. Comprehensive computational tests on HEAs demonstrate that GMHEA achieves a balance between accuracy and efficiency: the average computational time per atom ranges from 23.13 to 52.19 s. Structural analysis reveals that the configurations of HEAs exhibit short-range order (SRO) with optimized interatomic distances, correlating strongly with thermodynamic stability. This software provides a powerful tool for accelerating the design and development of HEAs by enabling rapid identification of stable atomic structures, with broad implications for applications in catalysis, energy storage, and advanced manufacturing.