“…Moreover, neural network models were used to predict the flow behaviors of 28CrMnMoV steel [19], glass fiber reinforced polymers [20], A356 aluminum alloy [21], and 42CrMo steel [22]. In addition, dynamic recrystallization (DRX) model [23], Zerilli-Amstrong (ZA) model [24,25], mechanical threshold stress plasticity (MTS) model [26], Cellular Automata (CA) model [27], and Bonder-Partom (BP) model [28] belong to the physics-based models. The study of V-4Cr-4Ti [29], medium carbon and vanadium microalloyed steels [30], and Mg-Al-Zn alloy [31] shows that the constitutive equation based on microscopic mechanism has good applicability, which can be used to characterize the relationship between flow stress and microstructure during high-temperature rheological process.…”