2050mm生产线精轧温度预报精度的提升

Prediction accuracy upgrade of Finishing rolling temperature of 2050mm rolling line

宝山钢铁股份有限公司2 050mm热轧线穿带速度设定功能有人工和自动两种方式,因自学习计算不合理导致遗传系数不合理而存在温度预报偏差大,自动速度投入率低等问题,通过优化温降计算参数和遗传学习策略,提高了温度预报准确性,从而提高了穿带速度的自动方式投入率。

The automatic control ratio of threading head speed in the 2 050 mm hot rolling line of Baoshan Iron Steel Co.,Ltd.was lower,and the temperature deviation between predicted and actual was bigger,because the unreasonable self-learning calculation resulted that the temperature heredity coefficient was unreasonable.Through optimizing the temperature drop mode and heredity learning strategy,the accuracy of temperature forecasting was improved,thus the automatic control ratio of threading head speed was increased.