氧化铝连续碳酸化解过程分解率专家控制系统

【作者】 卢宏燕；

【导师】 阳春华；

【作者基本信息】 中南大学 ， 控制理论与控制工程， 2005， 硕士

【摘要】 连续碳酸化分解过程是砂状氧化铝生产新工艺过程的关键工序,其分解率梯度与末槽分解率直接影响连续碳酸化分解过程产品的产量和质量。本文针对中州分公司连续碳酸化分解过程的控制现状,在分析影响氢氧化铝产品质量和产量因素的基础上,开发了基于专家经验的连续碳酸化分解过程分解率控制系统,实现了分解率的自动控制。系统已于2004年7月投入运行,实际运行情况表明,分解率合格率达到98%,槽样硅合格率达到95.2%,对提高Al（OH）₃产品的质量和产量作出了贡献。其主要工作和研究成果包括以下几个方面: 1.基于连续碳酸化分解机理及影响氢氧化铝产品质量和产量的因素,提出分解率专家控制的方案,即通过控制进料流量和4^#、5^#槽CO₂气体通入量来保证合适的分解率梯度和合格的末槽分解率。 2.考虑影响各槽分解率梯度的因素,建立分解率专家控制器。该控制器采用产生式规则进行知识表示,采用正向推理与深度优先搜索的方式进行推理。 3.建立末槽分解率自适应预估模型。该模型采用线性回归方法预估下一时刻的末槽分解率,通过其预估值对专家控制器的输出进行前馈补偿。 4.采用Visual C++开发了连续碳酸化分解过程优化控制软件。该软件实现了工艺参数的实时监视、进料流量和CO₂气体通入量的自动控制、故障的在线预测和报警等功能。更多还原

【Abstract】 The process of the continuous carbonation decomposition (PCCD) is the key procedure in the sandy alumina production, and the resolution ratio gradient and the last resolution ratio directly affect the product quality. Aimed at the existing condition of the PCCD and the factors to quality and quantity of the riopone product in the Zhongzhou Aluminum Group Company, a resolution expert control system of the PCCD is developed to realize the automatic control of the resolution ratio. The system has been in operation in July, 2004. The practical running results indicate that not only the qualified rate of the resolution ratio reaches 98%, and the qualified rate of Silicon reaches 95.2%, but also the product quality and quantity are improved. The following are the main jobs and the research findings in the paper.1. In order to ensure the optimal resolution ratio gradient and the last resolution ratio, it is necessary to adjust the feed flow and the aeration amount of the dioxide carbon in the forth and fifth decomposers in view of the mechanics of the continuous carbonation decomposition and the factors to the qualified riopone product.2. Taking into account the factors to the resolution ratio gradient, a resolution ratio expert controller is described, where the knowledge is represented in the form of the production rule, and the inference engine adopts the Forward chaining and Depth-first search.3. A self-adapting prediction model for the last resolution ratio is established. The model forecasts the next last resolution ratio by the method of linear regression. The output of the expert controller can be feedforward compensated by the next last resolution ratio.4. With Visual C++, an optimizing control system software for the PCCD is explored, which implements the functions such as the technological parameters monitoring, auto-control on the feed flow and the aeration amount, and on-line fault diagnosis and warning.更多还原