【作者】 王鑫；

【导师】 肖宏； 谢红飙；

【作者基本信息】 燕山大学 ， 机械设计及理论， 2009， 硕士

【摘要】 轧制压力预报是精轧机组计算机设定模型的核心,其预报精度将直接影响到辊缝的设定,穿带的稳定性,板厚精度以及产品的最终质量等等。这在日益激烈的市场竞争条件下,面对用户对钢材越来越高的要求,轧制压力预报就在轧制过程中显示出更加重要的地位。由于轧制过程中需要考虑相当多的线性和非线性的影响因素,非常复杂,就国内某钢厂不锈钢工程1780mm热轧生产线而言,采用的精轧设定模型中的轧制力预报模型精度并不是很理想,因此还需要进一步加以改进。本文采用离线模拟获得数据并对生产数据进行分析,系统的研究了热连轧带钢轧制力预报在线模型的建立方法,针对带钢中化学元素对带钢变形抗力的影响,用线性回归的数学分析方法对带钢中的各化学成分的变形抗力参数进行回归,调整了变形抗力模型,并通过轧制力自学习模型的分析用指数平滑法对修正后的轧制力计算进行自适应学习,提高了1780mm热连轧精轧轧制压力计算模型的预报精度和当钢种发生改变之后的适应性,不仅改善了轧制压力预报模型中变形抗力模型的计算精度,也使轧制力模型的自学习系数在钢种发生变化后仍然能保持一个平稳的趋势。总之,本课题在提高1780mm热连轧轧制压力预报模型预报精度和适应性方面的工作取得了初步的成效,但是由于限于理论研究,在数据量方面也不是很充足,因此仍然有待于扩充数据后进一步深入的研究与改进,并在此基础上实现轧制压力模型对更多钢种预报轧制力的良好适应性。更多还原

【Abstract】 The prediction for rolling pressure is the core of the finishing mill group computer model. Its forecast accuracy will directly affect the roll gap settings, the bite stability, thickness accuracy and finally the quality of the products and so on. In conditions of the increasingly competitive market, the higher users’requirements for steel, the more important position of the rolling forces prediction in the rolling process. Because we need to consider quite a number of linear and nonlinear factors in the rolling process, very complicated, as the result of the model using in the Stainless steel works 1780mm hot rolling production line of Bashan Iron Steel C o. is not very ideal, it is need to do a further improvement.In this paper, we use the offline simulation to obtain and then analysis the production data and study on the impact of the chemical element in the strip on the deformation resistance, and then return the deformation resistance parameters of the strip chemical composition in linear regression math analysis method. It optimizes the deformation resistance model, and through the analysis of the self-learning model of the rolling force. We have made optimized calculation of rolling force self-adaptive learning in exponential smoothing method, which improved the forecast accuracy of the 1780mm hot rolling mill finishing rolling pressure calculation model and the adaptability after its change. Not only improved the deformation resistance model calculation accuracy in the rolling force prediction model, but also the self-learning coefficient of the rolling force model changing in the steel can still maintain a steady trend.Anyway, in regard the 1780mm hot strip mill rolling force prediction model forecast accuracy and adaptability, this issue has achieved preliminary results. However, because of limited theoretical research, and the data volume is not adequate, the expansion of the data remains to be further research and improvement in-depth. And on this basis implement the good adaptability of the rolling pressure model for more steel rolling force prediction.更多还原