【作者】 马占福；

【导师】 赵西成； 李子文；

【作者基本信息】 西安建筑科技大学 ， 材料学， 2011， 博士

【摘要】 热轧带钢的板形控制是板带材生产领域的重点和核心,本论文依托1750热连轧机,以提高热轧薄板的板形质量为目标,深入地分析了板形控制基础理论和模型仿真,通过现场试验和模型仿真程序的开发,结合生产现场控制系统对板形控制的工艺和设备进行了研究,主要研究内容及结果如下:从模型结构和算法方面对板形控制理论进行了研究,结合生产现场系统地分析和阐明了影响板形控制的工艺和设备,通过对上下辊热凸度啮合对比试验,分析了在轧制过程中热凸度的变化过程,提出并实施了对部分冷却设备的改造,工作辊冷却水系统的改造,成功解决了改造前精轧F1～F3机架消凸过大的问题,保持了轧制稳定性,减少了中浪的出现,较为良好地控制了热轧带钢板凸度;提出了精轧支撑辊也使用CVC辊形技术的思路,使之与工作辊辊形配合,降低了轧辊的辊耗和剥落事故,在轧制相同吨位的条件下,换下辊后支撑辊CVC辊形比平辊形的磨损量降低了32.7%,延长了支撑辊的使用寿命,改善了板形质量,使板形控制有了显著提高。以Matlab软件为编程平台,利用人工神经网络技术,分别开发出弯辊力预报仿真模型和轧制力预报仿真模型,通过程序设计及模型计算,结果表明,利用人工神经网络模型进行仿真预报,其精度比生产现场的传统预报方式预报的精度有较大幅度的提高,弯辊力模拟预报值精度比传统的预报值提高了78.04%,轧制力预报的精度比传统的精度提高了36.71%,所建模型高精度地模拟逼近了实际值。CVC辊型曲线是影响板形控制的重要因素之一,结合生产现场的轧辊曲线,利用Matlab强大的数学分析功能进行了CVC辊型曲线数学模型的拟合求解,得到了更为合理的辊型曲线数学模型:y = 4.0326×10^-2-2.045×10^-3x+1.4522×10^-6x²-3.8045×10^-10x³此辊型曲线数学模型的建立进一步改进和优化了现场实际应用的辊型,使生产过程中的带钢板形控制有了显著提高。通过生产现场大量数据的收集及统计,分析了工作辊、支撑辊的磨损对板凸度控制产生的影响,通过轧辊磨损试验,分析出了在一个轧制周期内工作辊的磨损程度及规律,上工作辊磨损量大于相应下工作辊磨损量,下支撑辊磨损量大于相应上支撑辊磨损量;工作辊部分实际的磨损曲线基本与实测数据较相似,所获得的实测数据与轧辊磨损一般规律相一致,磨损量的大小与轧辊的使用期内所轧带钢长度、温度、钢种以及金属前滑、后滑、横向流动等因素有关,在热轧带钢轧机中,各机架的轧辊研磨和疲劳磨损幅度的总体趋势与轧制工艺参数相对应,使现场辊耗的控制、合理用运轧辊策略及优化轧辊磨削等技术得到了有效提升。板形控制中轧制规程的优化设计及实践应用,用动态规划的方法对热轧现有的轧制规程进行负荷分配优化计算,对轧制规程中的工艺参数进行了优化计算,由试验结果表明用动态规划的方法进行轧制规程的优化,对于良好板形的控制显示出了较强的优势,当来料出现厚度波动或其他扰动时可以对轧制规程进行适当调整,确立最佳的轧制规程,以此提高凸度与平直度的命中率。通过负荷分配的优化计算,合理分配轧机的道次压下量,使优化负荷分配的总功率比原现场使用的负荷分配的总功率平均降低了3.37%,进一步降低了生产负荷,由于轧制负荷的降低,减轻了轧辊凸度的磨损程度,辊凸度磨损变化幅度的降低使生产过程中带钢的板形质量得到了有效改善。由此本方案的实施优化了热轧厂原轧制规程,降低了轧机的负荷,改善了板形。更多还原

【Abstract】 Shape control of hot rolling strip is the focal point and key problem in the field of strip rolling. The model of shape controlling has been numerically and experimentally investigated and the basic theory of shape controlling is deeply analyzed by the equipment of 1750 hot strip mill for the purpose of improving the quality of hot rolling strip. The technology and equipment of shape control have been researched by field test and simulation program.The main research contents and the results are as follows:Shape controlling theory is studied by means of the model structure and algorithm. Technology and equipment influence factors of the shape control in the production field have been analyzed. The changing process of the thermal crown joggle between upper and bottom work rolls is analyzed by the contrast test. Several improvements were presented and water supply system of work roll has been reformed, which solved successfully the problem of the crown abrasion excessively among ahead three stands, keeping the steady rolling, decreasing the middle wave, and getting good quality of crown. The technique of CVC roll shape for finish back roll is proposed, which matches up work roll shape. So, the results show that the back roll failure accident decreases and the abrasion degree reduces 32.7% by contrast with flat back roll shape, which also prolongs the back roll service time and improves the shape quality.The simulation models of the bend force （BF） and rolling force （RF） prediction have been developed by the program platform of Matlab software and the technology of artificial neural networks （ANN）.By this way, the results present that the precision of model prediction is improved by a large margin by using of ANN instead of traditional model prediction way and the prediction accuracy of bend force is improved 78.04%, and prediction accuracy of roll force is improved 36.71%, and this model accurate prediction of ANN is highly close to the actual results in hot rolling .One of the important factors of influencing strip shape control is the CVC roller model curve. Combined with the actual roller curve in production field, the equation of CVC roller model curve is computed and got fit curve in arithmetic model:y = 4.0326×10^-2-2.045×10^-3x+1.4522×10^-6x²-3.8045×10^-10x³The equation of the arithmetic model improves and optimizes the roll shape curve and improves remarkably shape controlling in the process of producing.The effect of the work roll and back roll abrasion on the crown controlling of strip was analyzed by collecting a lot of data from the spot. By roller abrasion testing, the abrasion law and degree of work roll in a rolling period time has been got. The abrasion quantity of upper work roll is more than that of bottom work roll, and the abrasion quantity of bottom back roll is more than that of upper back roll. The gross trend of the roller grinding and fatigue abrasion in all rolling machines is coincidence with that of rolling process. The abrasion quantity is related with the strip coil length, temperature, grade of steel, slip forward, slip back, cross flow, etc. In hot mill, the roll grinding and fatigue abrasion trend is similar to the actual technological parameters, which decreases the roll consuming and improves the technology of roll strategy management and roll grinding optimization.The load assignment of the rolling schedule and the technology parameters have been optimized to design and apply to the rolling schedule in the process of shape controlling and calculated by dynamic programming. The test results show that the gross power of load assignment has the superiority by dynamic programming in the case of the thickness fluctuation or other disturbance. Optimal rolling schedule can be got by coordinating the original one, and the flatness or straightness accuracy is improved. After redistributing load assignment and adjusting the thickness value of 6 stands, the gross power averagely decreases 3.37% by contrast with original condition, and decreases the load, and alleviates the abrasion degree. As the result, the shape controlling quality is improved effectively. This scheme is applied in hot strip mill to reduce the load of the rolling mill and improve the whole stands shape controlling ability.更多还原