基于专家评估和信息融合的高炉料面温度场智能建模  

Intelligent modeling based on expert evaluation and inosculated information for burden surface temperature field of blast furnace surface

作　　者：安剑奇[1] 吴敏[1] 何勇[1] 曹卫华[1] 

机构地区：[1]中南大学信息科学与工程学院,湖南长沙410083

出　　处：《计算机与应用化学》2008年第7期782-786,共5页Computers and Applied Chemistry

基　　金：国家863计划资助项目(2007AA04Z177);国家杰出青年科学基金项目(60425310).

摘　　要：高炉料面温度分布对于高炉操作有着重要的指导意义。本文针对高炉目前无法准确全面了解料面温度分布的情况,挖掘现场已有数据,依据可信度理论,建立红外图像可信度评估的专家规则;根据评估结果,分别采用依靠红外图像的机理模型信息融合方法和不依靠红外图像的神经网络信息融合方法,建立基于专家评估和多源信息的高炉料面温度场模型。在某钢铁企业2 200 m^3高炉应用结果表明,该模型可以更为准确,直观地反映料面温度分布,实现了对料面温度分布的实时检测,为复杂冶金过程控制和监控提供了有效的解决方法。It is important to the operators of blast furnace to find out the distribution of burden surface temperature. In this paper, aiming at making clear the distribution of the burden surface overall and accurately, the expert rules are established to evaluate credibility of the infrared image by excavating the existed processing data and credibility theory. Based on the different result of evaluation, the surface temperature field of blast furnace model is established by the mechanism model inosculated information which depend on the infrared images or neural network inosculated information which do not depend on the infrared images. The application of the model in a 2 200 m^3 blast furnace indicates that the model can reflect the distribution of the burden surface temperature accurately and directly, thus the burden surface can be detected real time. This paper provides an effective solution for complicated metallurgy process controlling and monitoring.

关 键 词：高炉 料面温度场 红外图像 专家规则 信息融合 

分 类 号：O6[理学—化学] TP391[自动化与计算机技术—计算机应用技术]