Optimization of process parameters during synthesis of high entropy shape memory alloys through response surface and evolutionary method

High-entropy shape memory alloys (HESMAs) are considered promising materials for advanced engineering applications due to their unique properties, including the shape-memory effect, corrosion resistance, superelasticity, bioadaptability, and wear resistance. This study presents a modeling-based approach to optimize powder metallurgy process parameters for HESMA synthesis using a hybrid framework that integrates Response Surface Methodology (RSM), Box–Behnken Design (BBD), Artificial Neural Networks (ANN), and Genetic Algorithms (GA). The optimization processes were designed through RSM to evaluate the influence of key process factors, with theoretical density calculated via the Rule of Mixture serving as the primary response variable. The RSM model achieved a high coefficient of determination (R 2 = 0.9660), while the ANN–GA model demonstrated superior predictive accuracy with R 2 values of 0.97858 (training), 0.96096 (testing), and 0.98611 (validation). The overall regression coefficient (R 2 = 0.9693) confirmed the improved fit of the ANN–GA framework compared to conventional RSM. These results highlight the capability of the hybrid ANN–GA approach to model complex, nonlinear relationships between process parameters and theoretical density, thereby providing a systematic and data-driven methodology for process optimization in HESMA synthesis.