预测控制在水泥生料质量控制中的应用

发布时间：2018-05-20 19:50

本文选题：水泥生料质量控制 + 成分检测　；参考：《济南大学》2017年硕士论文

【摘要】：水泥是重要的建筑材料,水泥生料是生产水泥产品的基础,对水泥产品的质量影响甚大。国内大部分水泥企业依靠人工调节的方法生产水泥生料,这种调节方法准确性低、实时性差,难以得到高质量的水泥生料;实现水泥生料生产的自动化,提高水泥产品的质量是一个亟待解决的重要课题。近红外在线分析仪能够实时检测水泥生料中各种氧化物的含量,有助于实现水泥生料生产的自动化。但水泥生料生产过程具有滞后、惯性、扰动多等特点,难以控制,研究水泥生料的质量控制具有很大的意义。本课题以山东省内某水泥有限公司5000t/d的水泥产品生产线为研究背景,依据水泥生料的生产工艺,结合近红外在线分析仪及生产现场的情况,提出运用基于动态模型的预测控制调节方案和基于数学规划关系式的修正质量控制方案分别对不同的工况进行调节,改进水泥生料质量控制软件,主要内容如下:(1)水泥生料质量预测控制模块的开发。当水泥生料三率值中的石灰饱和系数需要调整而另外两个率值无需调整时,此时只需改变原材料石灰石和泥岩的配比便可达到调节目的。这种情况下原材料的配比只改变两种,水泥生料中各氧化物的含量也主要改变两种;由于本课题所在的水泥厂对石灰饱和系数非常重视,而且企业所用的石灰石化学成分不稳定,需要频繁的对石灰石配比进行调节,而且运用数学规划关系式的调节效果不够精准,但此种情况便于动态模型的辨识与使用。对此情况,本文将原材料配比的变化量作为输入量,水泥生料中CaO和SiO2含量的变化量作为输出量进行模型辨识,并对模型进行验证;然后根据模型设计相应的预测控制器,通过仿真确定必要参数。(2)水泥生料质量修正控制模块的开发。当水泥生料三率值中的铁率或硅率不合格时,需要调节原材料砂岩或铁尾渣的配比才能达到调节目的。此种情况下输入变量、输出变量数量较多且变量间的耦合性很强,精准的动态模型难以建立;在本课题所依托的水泥企业中,铁率或硅率一直比较稳定,调节频率很低,而且原材料砂岩和铁尾渣的化学成分稳定,通过数学规划关系式便可确定配比变化量与生料中各氧化物含量的变化关系。本文运用数学规划关系式结合专家规则对此种情况进行控制,算是对生料质量控制系统的修正控制。另外本文还提出基于最小二乘法原理对原材料成分进行修正的方法。最后通过对控制方法仿真,得到控制效果良好的参数。(3)水泥生料质量控制系统的实现。生产现场中,系统依据在线分析仪对实时数据进行采集,通过OPC实现硬件系统与数据库间的数据交互,利用水泥生料质量控制软件识别各种工况并对相应的工况进行调节。该系统能够实现水泥生料质量控制的自动化,提高水泥生料的质量。文中对阶段性的生产数据进行了统计与分析,分析结果表明水泥生料质量控制系统能够长时间稳定自动运行且控制效果优于人工控制效果。
[Abstract]:Cement is an important building material. The cement raw material is the foundation of the production of cement products. It has a great influence on the quality of the cement products. Most of the cement enterprises in China rely on manual regulation to produce cement raw material. This adjustment method is low in accuracy, poor in real time and difficult to get high quality cement raw material, and to realize the automatic production of cement. It is an important issue to improve the quality of cement products. The near infrared on-line analyzer can detect the content of various oxides in the cement raw material in real time, and help to realize the automation of the production of cement raw material. However, the production process of cement raw material has the characteristics of lag, inertia and many disturbances, so it is difficult to control the quality of cement raw material. Quantity control is of great significance. Based on the cement production line of a India Cements Limited 5000t/d in Shandong Province as the research background, based on the production technology of the raw material of cement, combined with the near infrared on-line analyzer and the situation of the production site, this paper puts forward the application of the dynamic model based pre test control and control scheme and the mathematical programming relationship. The modified quality control scheme is adjusted to different working conditions to improve the quality control software of cement raw material. The main contents are as follows: (1) the development of the prediction control module of cement raw material quality. When the lime saturation coefficient in the three ratio of cement raw material needs to be adjusted and the other two rate values need not be adjusted, only the raw lime should be changed at this time. The ratio of stone and mudstone can be adjusted. In this case, only two kinds of raw materials are changed, and the content of each oxide in the raw cement is changed mainly by two. Because the cement plant in this project attaches great importance to the lime saturation coefficient, and the chemical composition of the limestone used by the enterprise is unstable, and the limestone needs to be frequent to the limestone. The ratio is adjusted, and the effect of the mathematical programming relation is not accurate enough, but this situation is convenient for the identification and use of the dynamic model. In this case, this paper identifies the change amount of the raw material ratio as the input, the amount of CaO and SiO2 content in the raw material of cement as the output model identification, and verifies the model. Then the corresponding predictive controller is designed according to the model, and the necessary parameters are determined by simulation. (2) the development of the cement raw material quality correction control module. When the iron rate or the silicon rate in the three rate of cement raw material is not qualified, the ratio of the raw material sandstone or the iron tailings should be adjusted to achieve the adjustment purpose. In the cement enterprise relying on this project, the rate of iron or silicon is stable, the frequency of adjustment is very low, and the chemical composition of the raw material sandstone and iron tail slag is stable, and the proportion change and the raw material can be determined by the mathematical rule relation. This paper uses mathematical programming relations and expert rules to control this situation, which is the correction control of the quality control system of raw material. In addition, this paper also puts forward a method based on the principle of least square method to modify the composition of raw materials. Finally, the control effect is obtained by simulation of the control method. Good parameters. (3) realization of the quality control system of cement raw material. In the production site, the system collects real-time data according to the on-line analyzer, realizes the interaction between the hardware system and the database through OPC, and uses the cement raw material quality control software to identify various working conditions and adjust the corresponding working conditions. The system can be realized. The quality control of cement raw material is automated to improve the quality of cement raw material. The statistics and analysis of the stage production data are carried out in this paper. The results show that the quality control system of the cement raw material can run for a long time and automatically and the control effect is better than the artificial control effect.
【学位授予单位】：济南大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TQ172.4;TP273

【参考文献】