【作者】 张永辉；

【导师】 邵诚；

【作者基本信息】 大连理工大学 ， 控制理论与控制工程， 2006， 博士

【摘要】 铝粉氮气雾化生产工艺具有细粉率高，铝粉活性好，球形铝粉成型率高，生产安全等优点，为很多企业所采用。铝粉氮气雾化和分级工艺过程是一个复杂的工业过程，包括很多物理变化、化学反应以及一些动力学过程，具有强烈的非线性、强耦合、大滞后和不确定性等特点。生产过程中的关键控制变量雾化炉铝液温度用常规控制方法效果很差，直接影响了铝粉的细粉率和产量。作为生产质量指标的铝粉粒度分布无法直接实时地测量，给过程建模和优化控制带来了极大的困难。目前铝粉氮气雾化分级过程控制技术落后、自动化水平很低，过程优化控制问题成为提高氮气雾化铝粉细粉率和产品质量的一个瓶颈。本文以吉化集团高碳醇厂的铝粉生产装置改造项目为背景，在对铝粉氮气雾化和分级过程的机理深入分析的基础上，提出利用集成优化控制技术来实现铝粉氮气雾化分级过程的实时有效的控制和过程优化，并研制一套铝粉氮气雾化分级过程集成优化控制系统并成功应用于工业现场。本文的研究内容如下： 1．在分析铝粉氮气雾化分级过程机理特性和工艺流程特点的基础上，提出了以铝粉粒度分布为目标的铝粉氮气雾化分级过程集成优化控制系统的整体方案设计，采用IPC工控机+PLC+现场总线I／O模块的形式，利用工业以太网技术，组成三层分布式控制系统，实现了铝粉雾化、氮气循环、铝粉分级等全过程的集成优化控制，保证了生产过程的安全运行，减少了氮气和燃料气的消耗，应用实施结果验证了设计方案的有效性。 2．由于雾化炉铝液温度受许多不确定因素的影响波动很大，具有很强的非线性和大滞后等特性，难以建立确切的数学模型，其控制问题很难利用传统的控制理论和方法解决。本文针对雾化炉的铝液温度变化特性，结合模糊控制器和PID控制器的特点，设计了一种模糊-PID复合控制器，利用加权因子，将模糊控制器的输出和PID控制器的输出加权综合，使得控制器在误差较大时，主要由模糊控制器起作用，具有较快的响应能力，而在误差较小时主要由PID控制器起作用，具有较高的控制精度，实现了模糊控制器和PID控制器输出的连续平滑切换，在现场应用中取得了较好的控制效果。 3．铝粉氮气雾化生产过程中，作为质量指标和控制目标的铝粉粒度分布很难用传统的方法在线实时测量。本文提出利用RBF神经网络，通过测量与铝粉粒度分布有直接影响的其它变量，建立起铝粉粒度分布的软测量模型，实现了铝粉粒度分布的实时在线预测，为实施铝粉雾化过程优化控制提供了建模基础。 4．铝粉氮气雾化生产过程中，工况条件是不断变化的，根据专家经验得出的各工艺参数设定点不一定能保证雾化效果达到最优，因此需要对铝粉氮气雾化过程实施优化更多还原

【Abstract】 For the nitrogen atomizing technology is characterized by higher fine powder proportion, the reactive feature of aluminum powder, higher spherical shaped proportion of aluminum powder, safer producing process and many other advantages, the production technics has been widely adopted by many aluminum powder enterprises. The process of aluminum powder nitrogen atomizing and classifying is a complex industrial process. It includes many physical and chemical reactions and some dynamics processes, and it has characteristics of strong nonlinear, strong coupling, large lag, uncertainty and so on. The melted aluminum’s temperature in the atomization furnace is the key control variable of the atomizing process, which has a great effect on the fine powder proportion and the production of the aluminum powder. This temperature cannot be controlled well by the conventional control methods. As the production quality index, the distribution of the particle size of the aluminum powder cannot be measured directly, and it brings great difficulties to the process modeling and control optimization. At present, for the aluminum powder nitrogen atomizing and classifying process control technology is outdated and the level of automation is low, the optimal control technology turns into the important method of improving aluminum power quality. In this paper, by taking the reconstruction of the aluminum powder production lines in High Alcohol Plant of CNPC Jilin Chemical Group as the backdrop, the mechanism of the aluminum powder nitrogen atomizing and classifying processes is analyzed comparatively and deeply. The author proposes the aluminum powder nitrogen atomizing and classifying process intelligent integrated optimal control system for the first time, and has put it into use successfully. The content of the research in this paper is as follows:1. Based on the analysis of the mechanics of the aluminum powder nitrogen atomizing process and the technological characteristics, taking aluminum powder particle size distribution as the optimal control objective, the project of the integrated optimal control system of the aluminum powder nitrogen atomizing and classifying process are proposed. In this project, the 3-layer distributed control system is presented by assembling PLC, industrial PC, field bus I/O modules and the industrial ethernet, and realizes the integrated optimal control of the aluminum powder nitrogen atomizing and classifying process. The application of this integrated optimal control system guarantees safe operation of aluminum powder更多还原