Changsha University

High-entropy alloys (HEAs) are emerging as next-generation structural materials due to their outstanding mechanical and functional properties. However, their vast compositional and configurational complexity poses major challenges for conventional trial-and-error approaches and ab initio simulations, which struggle with high computational costs and limited predictive accuracy. Existing machine learning approaches, while promising, remain constrained by data scarcity, limited interpretability, and the lack of effective human-AI interaction. To address these limitations, we introduce an integrated human-computer interactive HEA design platform that incorporates emotional feedback. By combining natural language processing (T5 model), machine learning (XGBoost), and multi-objective optimization (NSGA-II), the platform establishes a closed-loop "perception-decision-optimization" workflow. Real-time emotion recognition dynamically adjusts optimization weights, enabling efficient human-AI collaboration. The model achieves high accuracy in predicting yield strength and Young's modulus, with SHAP analysis revealing the underlying physical mechanisms. Emotion-driven optimization guides Pareto front convergence, with results showing <1.4% deviation from experimental values. This high degree of accuracy underscores the efficacy of integrating affective feedback into the optimization loop, enabling a more responsive and user-aligned design process that effectively bridges subjective expert preferences with quantitative multi-objective optimization. The multiscale modeling further validates the platform's reliability for complex dual-phase alloys. This work establishes a novel paradigm for interpretable and efficient AI-driven material design, highlighting the transformative potential of integrating artificial intelligence with expert knowledge.

Abstract Achieving high-precision imaging of hidden cracks in tunnel linings is challenging due to strong interference from rebar clutter. To address this issue, we propose a collaborative imaging method that integrates Edge-Enhanced High-Resolution Reverse Time Migration (EEHR-RTM) with Adaptive Frequency-Wavenumber Migration (A-FKM). The EEHR-RTM method accurately reconstructs the geometric contours of both rebars and cracks, while the A-FKM framework effectively suppresses strong rebar interference signals and preserves the amplitude of defect reflections. Validation through FDTD simulations and laboratory model tests demonstrates that, compared to conventional RTM, the proposed method significantly improves key quantitative metrics: the Signal-to-Noise Ratio increases from 5.40 dB to 29.58 dB, the Gradient Energy increases from 2.417×10⁹ to 1.110×10¹⁸, and the Edge Density increases from 0.004 to 0.027. Furthermore, compared to traditional F-K migration, LRSD, and SVD, the proposed method achieves a maximum rebar clutter removal rate of up to 87.9%. This approach enables high-definition imaging of millimeter-scale cracks even under strong rebar interference, providing a reliable technical pathway for the nondestructive testing of hidden defects in tunnel linings.

Abstract A 54‐day feeding trial was conducted to investigate the effects of dietary protein levels on growth performance, digestive capacity, plasma biochemical indexes, and gene expression in Hefang crucian carp ( Carassius auratus ). Six experimental diets were prepared with protein levels of 21.7%, 24.0%, 29.4%, 34.0%, 38.0%, and 43.2%, respectively. Each diet treatment was randomly assigned to triplicate groups of 25 fish, with an initial body weight of (14.60 ± 0.39) g per tank. Fish were fed twice daily to apparent satiation. The results showed that weight gain rate (WGR) and specific growth rate (SGR) exhibited an upward trend with increasing dietary protein levels, and reached the highest value in the 43.2% protein group. Conversely, protein efficiency was significantly higher in the 24.0% protein group compared to the other protein groups ( p &lt; 0.05). As dietary protein levels increased, there was a significant rise in the condition factor (CF), while both the hepatopancreas somatic index (HSI), viscerosomatic index (VSI) and crude lipid of whole body experienced significant decreases ( p &lt; 0.05). The intestinal trypsin activity was significantly elevated in the high‐protein groups (38.0% and 43.2% protein) compared to the low‐protein group (21.7% protein), whereas the plasma triglyceride concentrations exhibited a significant inverse relationship ( p &lt; 0.05). However, at a dietary protein level of 43.2%, the lipid droplet content in the liver of Hefang crucian carp was significantly elevated compared to other experimental groups. Additionally, the group fed the diet with a protein level of 21.7% exhibited significantly higher lipid droplet content than the groups with protein levels of 24.0% and 34.0%. A curvilinear regression analysis indicated that the lipid droplet area in the liver was minimized at a dietary protein concentration of 29.9%. Regarding intestinal gene expression, the asct2 and lat2 genes exhibited the highest expression levels in the group supplemented with 43.2% protein. Conversely, the expression levels of pept1 and cdx2 genes initially decreased and subsequently increased as the dietary protein levels rose, achieving their peak in the low‐protein groups (21.7% and 24.0% protein) ( p &lt; 0.05). Concerning hepatopancreas genes, the expression levels of tor , igf1 , and gh genes were significantly elevated in the 24.0% protein group compared to other treatment groups. Additionally, the expression levels of ghr were significantly higher in the 21.7% and 43.2% protein groups compared to the other treatment groups ( p &lt; 0.05). In conclusion, Hefang crucian carp exhibited optimal growth performance at a dietary protein level of 43.2%, whereas a protein concentration of 29.9% was identified as the critical threshold for maintaining optimal liver health.