Intelligent prediction model for mechanical properties of cold-rolled strip steel based on bio-inspired regulation mechanisms and operational decision-making theory

To address the challenges in predicting the mechanical properties of cold-rolled strip steel posed by multi-process parameter coupling, system nonlinearity, and time-varying characteristics, this paper proposes an intelligent prediction model that integrates bio-inspired regulation mechanisms and operational decision-making theory. First, based on the dynamic feedback mechanism of the biological endocrine system, the model constructs an information feedback structure with self-adaptive regulation capability. Then, by incorporating the multi-objective priority factor theory from operations research, a dynamic allocation strategy for feedback signal weights is designed to optimize the model's response performance under complex working conditions. Furthermore, the hormone-inspired regulation mechanism of the biological endocrine system is introduced to realize real-time self-correction of model parameters during training, effectively mitigating prediction deviations caused by system dynamics. Experimental validation based on actual production line data shows that, compared to long short-term memory and least squares prediction models, the proposed model delivers the best overall performance in predicting tensile strength, yield strength, and elongation after fracture. It achieves the smallest root mean square error, mean absolute error, and mean absolute percentage error, with all coefficients of determination exceeding 0.99. Ablation experiments confirm the necessity of the biological feedback mechanism and parameter dynamic updating modules. This model can perform high-precision intelligent prediction of the key mechanical properties of cold-rolled strip steel, and holds significant technical support value for improving the precision of production quality control, optimizing the presetting of process parameters, and ensuring the stability of product performance.