论文部分内容阅读
针对国内热轧生产线上精轧区测温点稀疏、温度精度要求高的特点,采用二维有限差分方法建立了精轧区板坯温度的机理模型,通过现场实测数据并结合粒子群算法对温度模型进行优化。针对轧制过程中影响因素多、环境复杂的特点,将支持向量机引入精轧温度模型的预测,并与机理模型和实测数据对比,检测其有效性。经过验证,所建模型的计算精度与实测数据误差在10℃以内。利用所建模型,对终轧区喷水量和加速度提出一种综合控制策略,可为实际生产提供参考。
In view of the sparse temperature measurement point and the high temperature accuracy requirement in the finishing rolling zone in the domestic hot rolling production line, a two-dimensional finite difference method was used to establish the mechanism model of the slab temperature in the finishing zone. Based on field measurements and particle swarm optimization, Model optimization. Aiming at the characteristics of many factors influencing the rolling process and complicated environment, the support vector machine is introduced into the prediction of the finish rolling temperature model. Compared with the mechanism model and the measured data, the effectiveness of the support vector machine is tested. After verification, the error between the calculated accuracy of the model and the measured data is within 10 ℃. Using the model, a comprehensive control strategy is proposed for the water spray and acceleration in the final rolling zone, which can provide a reference for actual production.