基于工况参数预测模型的铅锌烧结过程操作优化系统设计

【作者】 岑丽辉；

【导师】 吴敏；

【作者基本信息】 中南大学 ， 控制理论与控制工程， 2003， 硕士

【摘要】 铅锌烧结过程是一个极其复杂的生产过程，其优化控制模型及相应软件系统的研究开发工作起步较晚，使得现场操作缺乏科学的指导依据。为此进行烧结过程工况预测与操作优化系统的研究，开发具有自主版权的建模和优化软件，将现有的人工智能等领域的重要成果应用于铅锌烧结过程，对优化铅锌烧结生产，实现生产过程综合自动化具有重要意义。 本文以韶关冶炼厂铅锌烧结过程为背景，介绍了铅锌烧结过程工况预测与操作优化系统的研究和设计方法。首先在基于机理分析的基础上，分析烧结过程中存在的问题，并且得到影响烧结状况的关键因素；然后提出解决方案，将监视及预报系统分为工况综合评判模块和操作参数优化模块。然后分别讨论了几个模块的建立过程及方法，包括基于BP神经网络的烧结综合透气性预报模型、结合主元分析和神经网络的烧结块产量、含铅量、含硫量的神经网络模型、结合综合透气性和烧结块产量质量预测模型的工况综合判断集成模型，以及基于工况综合判断模型的聚类分析参数优化算法；最后，讨论了系统的结构、功能，并进行了系统软硬件设计，使系统得到实现，这里重点突出了系统软件的设计思想和设计方法。 系统软件采用vc编制，主要包括了通讯模块、主监视模块、综合透气性模块、烧结块产量质量模块、工况综合判断模块等。系统通过现场调试运行，实现了烧结过程参数监视，综合透气性，烧结块小时产量、烧结块含铅量和烧结块含硫量预测，及操作参数优化等功能。现场运行情况证实系统具有一定的可靠性和实用性，能够满足现场要求的精度。更多还原

【Abstract】 The Lead & Zinc sintering process is a quite complex process. The optimal control model and software are not fully developed at present, which make it hard to give scientific directions to the fieldwork. Therefore, an Operating Optimization System is designed for this target, which is significant for optimizing production of this sintering process by applying artificial intelligence.Firstly the main problems of this sintering process are proposed and the key factors that influence the process are acquired by the analysis of mechanism. Secondly the solution to these problems is put forward. This Operating Optimization System is divided into a comprehensive evaluation module for the conditions and a parameter optimizing module. These two modules include a comprehensive permeability prediction model, a prediction model of agglomerate quality and quantity, an intelligent integrated prediction model for the conditions and a parameter optimizing algorithm which are realized by a combination of the Back-Propagation Network, principle component analysis, fuzzy comprehensive evaluation and cluster analysis.At last the system is expatiated in its structures and functions, which is implemented by software and hardware design. Here, the thoughts and the methods of the system design are emphasized. The system software, developed by Visual C++ 6.0, comprises a main monitoring module, a communication module and database module. An real-world application in the Shaoguan Smelt shows that all the functions are realized. The prcision and reliability have been proved.更多还原