论文部分内容阅读
基于人工神经网络建立了反向凝固过程中的性能预测模型，实现了对铸带厚度和新相层晶粒度的全面预测；探讨了凝固过程中的主要工艺参数对上述性能的综合影响，为反向凝固性能的综合预测提供了简便的新手段．研究表明，新生相晶粒度随钢水过热度、母带厚度、浸入时间变化对其影响不显著，而钢水过热度、母带厚度、浸入时间变化对铸带厚度的影响较大．该模型的预测结果与实测的结果较为接近．
Based on the artificial neural network, the performance prediction model in the process of reverse solidification was established, and the overall prediction of the thickness of the belt and the grain size of the new phase layer was realized. The comprehensive effects of the main process parameters in the solidification process on the above properties were discussed as Comprehensive prediction of reverse solidification provides a simple and convenient method. The results show that the grain size of the newborn phase has no significant effect on the thickness of the strip with the superheat degree of the molten steel, the thickness of the mother strip and the immersion time. However, the influence of the overheating degree of the molten steel, the thickness of the mother strip and the immersion time has a significant influence on the thickness of the strip. The model predictions and the measured results are more similar.