【作者】 李燕红； 纪建军； 田慧欣； 肖冬；

【Author】 Li Yanhong~1 Ji Jianjun~2 Tian Huixin~3 Xiao Dong~3 (1.BaoSteel Corporation,Shanghai,201900;2.Shanghai kings-tech Co.Ltd,Shanghai,200940;3.Information Science and Engineering,Northeastern University,Shenyang,110004)

【机构】 上海宝钢股份有限公司； 上海鼎实科技有限公司； 东北大学信息科学与工程学院；

【摘要】 本文针对软测量技术的特点,以及现有基于人工智能的软测量方法的不足,提出了一种基于信息融合ELM神经网络的软测量建模新方法。该方法将信息融合与ELM人工神经网络新方法相结合,有效地提高了软测量的精度。将该新方法应用到LF钢水温度软测量中,根据LF的冶炼特点分析出影响温度的主要因素,并将其作为模型的输入,建立软测量模型。使用实际生产数据对其进行检验,结果表明该基于信息融合ELM神经网络的LF钢水温度软测量方法具有较好的预测效果,完全可以满足实际生产的需要。更多还原

【Abstract】 Aiming at the characters of soft sensor and the disadvantages of the methods based on artificial intelligence,a new method which combining the information fusion and ELM is proposed to establish the model of soft sensor.This new method can enhance the accuracy efficiently.The new method is used to predict the temperature of molten steel in LF.According to the characters of LF refining,the main factors which affect the temperature are analysed,and are used as the inputs of model of soft sensor.The practical production data are used to test the soft sensor of temperature.The results demonstrate the new soft sensor of molten steel temperature in LF based on the new method can predict the temperature accurately.更多还原