基于无模型自适应的重介悬浮液密度控制

发布时间：2019-01-23 08:02

【摘要】：在选煤过程中,重介分选由于分选粒度范围宽、分选精度高、适应性强且易于实现自动化等优点,逐渐成为选煤的主要分选方式。作为重介选煤过程中最重要的影响因素,悬浮液密度的变化直接影响精煤产品的质量及分选效率,因此对悬浮液密度的控制尤为重要。悬浮液密度控制系统是一个高度非线性、时变、强耦合、大时滞的过程,很难建立被控对象的精确模型,传统的控制算法在控制速度及控制精度方面逐渐显现出了不足。同时,随着原煤性质等参数的变化,密度的设定值也必须随之改变以确保精煤产品符合要求。针对上述问题,本文分别从悬浮液密度控制和设定值预测两方面进行研究。论文首先介绍了重介选煤的工艺流程,并详细阐述了重介旋流器精选过程及其主要影响因素。在分析输入输出关系的基础上,根据试验采集到的数据,采用机理知识与试验数据相结合的混合法对重介悬浮液密度控制系统进行建模,并设计了重介悬浮液密度控制系统整体方案。在控制系统整体方案中,悬浮液密度的控制采用不依赖于数学模型的无模型自适应控制(MFAC)算法。在介绍MFAC基本理论知识的基础上,对悬浮液密度控制系统中的MFAC控制器进行设计与开发,并与常用的PID及模糊PID控制算法进行仿真对比。针对MFAC在滞后情况下存在的问题,设计了抗滞后MFAC(ADMFAC)控制器,并与标准MFAC在标称模型及滞后增加的情况下分别进行仿真对比。在MFAC算法中步长序列kr和权重因子l是影响系统性能的主要参数,然而,对于MFAC参数的整定尚缺乏完善的理论方法。为此,本文在分析蚁群算法基本原理、参数影响以及算法优缺点的基础上,选择使用最大最小蚂蚁系统(MMAS)对ADMFAC参数进行整定,得到步长序列kr和权重因子l的最优组合使ADMFAC控制性能达到最优。针对悬浮液密度设定值的预测问题,利用极限学习机(ELM)算法对预测模型进行训练。首先介绍了影响密度设定值的因素,然后简述了ELM算法的基本知识,建立了悬浮液密度设定值的预测模型,最后对建立的模型进行预测并与实际值对比。
[Abstract]:In the process of coal preparation, heavy medium separation has gradually become the main separation method because of the advantages of wide particle size range, high separation accuracy, strong adaptability and easy to realize automation. As the most important factor in the process of heavy medium coal preparation, the change of suspension density directly affects the quality and separation efficiency of coal concentrate products, so it is particularly important to control the density of suspensions. The suspension density control system is a highly nonlinear, time-varying, strongly coupled and time-delay process. It is difficult to establish an accurate model of the controlled object. At the same time, with the change of raw coal properties and other parameters, the density setting value must be changed to ensure that the clean coal products meet the requirements. In order to solve the above problems, the control of suspension density and the prediction of the set value are studied in this paper. In this paper, the process of heavy medium coal preparation is introduced, and the cleaning process of heavy medium cyclone and its main influencing factors are described in detail. Based on the analysis of the relationship between input and output, a hybrid method combining mechanism knowledge with experimental data is used to model the density control system of heavy media suspension. The whole scheme of density control system of heavy medium suspension is designed. In the overall scheme of the control system, the (MFAC) algorithm of model-free adaptive control is used to control the suspension density. On the basis of introducing the basic theoretical knowledge of MFAC, the design and development of MFAC controller in suspension density control system are carried out, and the simulation results are compared with the usual PID and fuzzy PID control algorithms. An anti-lag MFAC (ADMFAC) controller is designed to solve the problem of MFAC in the case of lag, and compared with the standard MFAC in the case of nominal model and the increase of lag. In the MFAC algorithm, step size sequence kr and weight factor l are the main parameters that affect the system performance. However, there is no perfect theoretical method for MFAC parameter tuning. Therefore, on the basis of analyzing the basic principle of ant colony algorithm, the influence of parameters and the advantages and disadvantages of the algorithm, this paper chooses to use the maximum and minimum ant system (MMAS) to adjust the ADMFAC parameters. The optimal combination of step size sequence kr and weight factor l is obtained to optimize the control performance of ADMFAC. Aiming at the prediction of suspension density, the prediction model is trained by (ELM) algorithm. This paper first introduces the factors influencing the density setting, then introduces the basic knowledge of the ELM algorithm, establishes the prediction model of the suspension density setting value, and finally forecasts the established model and compares it with the actual value.
【学位授予单位】：中国矿业大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：TD94

【相似文献】