基于多子系统信息熵的焦炉加热燃烧过程工况识别  

Operating-State Identification for Coke Ovens Combustion Process Based on Multi-Subsystem-Entropy

作　　者：雷琪[1,2,3] 蔡伟[1,3] 吴敏[1,3] 陈鑫[1,3] 

机构地区：[1]中南大学信息科学与工程学院,湖南长沙410083 [2]系统控制与信息处理教育部重点实验室,上海200240 [3]先进控制与智能自动化湖南省工程实验室,湖南长沙410083

出　　处：《信息与控制》2014年第3期361-367,共7页Information and Control

基　　金：国家自然科学基金资助项目(61203018);湖南省自然科学基金资助项目(11JJ6048);系统控制与信息处理教育部重点实验室开放基金(SCIP2011002);国家863计划课题(2012AA040307);中南大学博士后基金资助项目

摘　　要：针对焦炉生产过程的连续性和间歇性,提出一种基于多子系统信息熵的焦炉加热燃烧过程工况识别方法.首先从焦炉热平衡的角度对工况进行划分,结合焦炉结构和传热机理将焦炉分解为多个近似独立的加热子系统;然后基于对荒煤气温度的定性趋势分析判断炭化室结焦时段,分析各个加热子系统的耗热水平;其次提出一种采用信息熵确定各类子系统对焦炉不同工况影响权重的方法,实时判断焦炉加热燃烧过程工况.最后,仿真及应用证明了该方法的有效性.In view of continuity problems in coke production, a multi-subsystem entropy-based operating-state iden- tification is proposed for coke oven combustion processes. First, the operating states are divided according to heat balance. Then, the coke oven is decomposed into many subsystems based on structural characteristics and heat-transferring mechanism. On the basis of the trend analysis for raw gas temperature, the coking stage of coking chambers is judged, and the heat consumption level of each subsystem is then analyzed. An infor- mation-entropy-based method for evaluating the subsystems＇ influence on different operating-states is applied to identifying the operating state for the coke oven＇s combustion process. Finally, the effectiveness of the proposed method is verified by simulation and application.

关 键 词：焦炉 子系统 信息熵 定性趋势分析 工况识别 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]