烧结过程烧结终点智能控制策略及工业应用

【作者】 廖环宇；

【导师】 吴敏； 王春生；

【作者基本信息】 中南大学 ， 控制科学与工程， 2010， 硕士

【摘要】 烧结过程是钢铁冶炼的重要过程,烧结矿是高炉炼铁的主要原料,烧结矿的质量与产量直接影响到炼铁生产的质量与产量。烧结终点是烧结过程中最重要的热状态参数,是判断烧结过程正常与否的标志之一。然而由于烧结过程的高度复杂性和时滞时变性、部分参数难以检测、系统特性不确定、扰动因素众多等原因,烧结终点一直存在非常大的波动,且抗干扰性能很差。本文在分析烧结过程工艺机理的基础上,总结出了烧结终点控制问题中的主要难点。通过综合运用系统辨识与智能控制理论,提出了一种烧结过程烧结终点智能控制策略,实现烧结终点的稳定控制,使烧结矿质量与产量指标满足生产要求。首先,针对一些重要状态参数的不可检测性,分别建立烧结终点软测量模型及垂直烧结速度软测量模型。其次,基于烧结过程特性及烧结终点影响因素的分析,选择具有外界输入的自回归模型(ARX)结构描述烧结过程,采用闭环辨识算法得到烧结终点的预测模型。然后基于预测模型,设计前馈模糊控制器,适应水分、料层厚度等前馈扰动因素所造成的不稳定工况；并且基于垂直烧结速度的软测量模型,从机理上计算合适的台车速度,提高稳定工况下烧结终点的控制精度。实验仿真验证了本文所提建模方法与控制策略的有效性。同时为了验证其实际应用价值,针对国内某大型钢铁企业烧结厂,在原有自动化控制系统的基础上,开发烧结终点智能控制系统。实际运行结果表明：该系统实现了对烧结终点的控制,降低了烧结终点的波动率,减少了工人的劳动强度,有效地提高烧结机的利用系数,改善了烧结矿的质量。更多还原

【Abstract】 Sintering process is essential in the procedure of the iron and steel smelting, sinter, the quality of which has a direct impact on the quality of steel, is a vital raw material for blast furnace. Burn-through Point (BTP) is the most important heat state parameter in sintering process, and a sign to determine whether the process is normal or not. Sintering process is a complex system with time-dely and time-varying. BTP often fluctuates and cannot suppress the disturbance, due to the difficulty to detect some parameters, uncertainty and disturbance.This paper analysised the mechanism of sintering process, and summarized the main difficulties existed in BTP control problem. Based on the system identification and intellegent control theory, BTP intellegent control stategy is proposed to achieve the stability of BTP and ensure that the quality and quantity indicators meet the requirements.First, these two soft-sensoring models, including BTP and vertical sintering speed, are builded, due to some important parameters cannot be detected on-line. Through the analysis of the characteristics of sintering process and the influencing factors of BTP, an AutoRegressive with eXternal (ARX) model structure are defined to describe sintering process, and BTP prediction model is obtained by closed-loop identification method. Then a feed-forward fuzzy controller is designed to counteract the influence from some feedforward disturbance, such as water, bed height and so on. Besides, in order to improve the accuracy of BTP control in stable operation situation, the optimal stand speed is computed based on the current vertical sintering speed obtained from the soft-sensoring model.Simulations validate the effectiveness of the strategy proposed in this paper for modeling and controlling. Meanwhile, in order to verify its value of practical application, BTP intellegent control system is developed based on the original control system of a sintering plant in an iron and steel enterprise. The running results show that the system achieve the stable control of BTP, reduce the fluctuation and labor intensity, and effectively improve the utilization factor of the sintering machine and the quality of sinter.更多还原