基于分布参数模型的高炉炉温预测

Prediction of blast furnace temperature based on the distributed parameter model

高炉炉温的高低直接影响了高炉的生产稳定性以及炼铁过程中的低能耗、高质量等性能指标.本文采用分布参数系统思想建立了偏微分方程预测控制模型.首先,使用普通最小二乘方法分两种不同方案对模型进行参数估计并对他们的仿真结果进行比较分析,然后,使用易适应的参数估计方法(flexibleleast squaremethod,FLS)对模型进行变系数估计处理.此方法能够体现系数的时变性,提高模型的准确度.最后,从仿真效果图中展示了本方法对硅含量预报的结果无论在命中率上还是精确度上的优势和有效性.

Blast furnace temperature directly affects the stability of iron-making process,and other performance indicators such as low-power and high-quality.In this paper,distributed parameter systems are employed to build the predictive control model described by partial differential equations.First,different schemes are applied to estimate the parameters of BF model using ordinary least squares method.Additionally,the effects of the two schemes are compared based on the simulation results.Then,the flexible least square method （FLS） is used to estimate the time-varying coefficients of the mode.This method can reflect the time-varying coefficients and improve the accuracy of model.Finally,the simulation shows the method for prediction of silicon content is advantageous and effective both in the hit rate and in accuracy.