石化工业循环冷却水预测模型的研究及专家系统的开发

发布时间：2018-05-10 09:30

本文选题：工业循环冷却水 + 水质预测　；参考：《天津理工大学》2017年硕士论文

【摘要】：伴随节水减排的绿色发展要求,石化行业生产中普遍采用了循环用水模式,高浓缩循倍数下循环用水给整个生产运行提出更高要求,因此开展对工业循环冷却水腐蚀结垢的预测模型研究及专家系统的开发,该选题具有重要的理论意义和实际的经济价值。首先,工业循环冷却水系统正常运转中产生的腐蚀结垢问题对其影响巨大,造成严重的安全隐患。本文从中石化某分公司提取海量水质数据,经处理分析得出与腐蚀和结垢密切相关的六种指标,在此基础上利用NARX动态神经网络建立腐蚀速率和粘附速率的预测模型,并用Matlab进行大量的仿真验证,通过与其他神经网络的对比,证明建立的预测模型具有较为优异的预测性能。其次,针对传统的依靠人工经验投加药剂存在的弊端,本文设计开发了工业循环冷却水专家系统,并对关键的知识库和推理机做了较为详细的设计。之后基于C/S结构建立分层架构,利用可视化语言VB 6.0作为工业循环冷却水专家系统前端开发平台,采用MySQL作为后台系统数据库,同时针对VB在面对复杂数值计算和分析时的不足,利用VB与Matlab间的接口程序,把所需要的功能放到Matlab中实现。将建立的水质模型嵌入专家系统,使其不仅能够预测水质腐蚀速率和粘附速率预测,还可以提出合理可行的处理建议,经循环水系统运行检测,提高了实际效率,减小了预测周期,为冷却水系统安全运行提供了最大程度的保证。
[Abstract]:With the green development requirement of water saving and emission reduction, the circulating water mode is generally adopted in the production of petrochemical industry. The high concentration cycle water circulation water gives higher requirements to the whole production operation. Therefore, the prediction model of corrosion and scaling of the industrial circulating cooling water and the development of expert system are carried out. The topic has important theoretical significance. First, the problem of corrosion and scaling produced in the normal operation of the industrial circulating cooling water system has a huge impact on it, causing a serious safety hazard. This paper extracts massive water quality data from a petrochemical company in Sinopec, and obtains six indexes closely related to corrosion and scaling through processing and analysis. On this basis, the NARX dynamic is used. The prediction model of corrosion rate and adhesion rate is established by state neural network, and a lot of simulation verification is carried out with Matlab. By comparing with other neural networks, it is proved that the prediction model has excellent prediction performance. Secondly, in view of the disadvantages of traditional drug adding agent, this paper designed and developed industry. The expert system of circulating cooling water is designed in detail, and then the key knowledge base and reasoning machine are designed in more detail. Then a layered architecture is built based on the C/S structure. The visual language VB 6 is used as the front end development platform of the industrial circulating cooling water expert system, and MySQL is used as the data base of the background system, while the complex numerical calculation is faced to the VB in the face of the complex numerical calculation. And the shortage of analysis, using the interface program between VB and Matlab, put the required function into the Matlab. The water quality model is embedded in the expert system, so that it can not only predict the water quality corrosion rate and the prediction of the adhesion rate, but also put forward a reasonable and feasible treatment proposal. Efficiency, reduce the prediction period, and provide the greatest guarantee for the safe operation of the cooling water system.

【学位授予单位】：天津理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TQ085;TP18

【相似文献】