摘要
为克服炉缸局部热状态评判的局限性,全面考虑高炉各冷却壁段水温差对热状态的影响,该文建立基于时间序列的高炉水温差多维度模糊综合评判模型。首先,选取炉身、炉腰、炉腹、炉缸各冷却壁段水温差时间序列,用相关系数法确定上部各段水温差对下部水温差影响的滞后时间,并筛选相关系数大于阈值的参数为主要指标参数;其次,用概率统计法划分指标参数最优区间,计算其相对劣化度值,选择分段岭形分布函数计算各指标参数隶属度值;最后,对各层指标参数自下而上进行模糊合成,结合指标参数权重和最大隶属度原则综合获得最终评判等级。经实际数据仿真分析,模型评判命中率为81.7%,能实时地评判高炉实际水温差状态。
In order to overcome the limitation of the local thermal state evaluation of the hearth,the influence of the water temperature difference of each cooling section of the blast furnace on the thermal state is considered comprehensively.Based on the time series,the multi-dimensional fuzzy comprehensive evaluation model of the blast furnace water temperature difference is established.Firstly,the time series of water temperature difference between the cooling wall sections of the furnace body,the furnace waist,the belly and the hearth is selected.The correlation coefficient method is used to determine the lag time of the influence of the water temperature difference on the lower water temperature difference,and the parameters whose correlation coefficient is greater than the threshold are selected as the main indicator parameters.Secondly,the optimal interval of index parameters is divided by probability and statistics method,the relative deterioration degree value is calculated,and the segmentation ridge distribution function is selected to calculate the membership degree value of each index parameter.Finally,the fuzzy index synthesis of each layer index parameter is performed from bottom to top.The index parameter weight and the maximum membership degree principle are combined to obtain the final evaluation level.Through the actual data simulation analysis,the model evaluation hit rate is 81.7%,and the actual water temperature difference state of the blast furnace can be judged in real time.
作者
崔桂梅
张胜男
张勇
马祥
CUI Guimei;ZHANG Shengnan;ZHANG Yong;MA Xiang(School of Information Engineering,Inner Mongolia University of Science and Technology,Baotou 014010,China;Iron-making Plant,Inner Mongolia Baogang Steel Union Co.,Ltd.,Baotou 014010,China)
出处
《中国测试》
CAS
北大核心
2019年第9期13-20,共8页
China Measurement & Test
基金
国家自然科学基金(61763039)
关键词
多维度
时间序列
相关系数法
概率统计分析
相对劣化度
模糊综合评判
multi-dimensional
time series
correlation coefficient method
probability and statistical analysis
relative deterioration degree
fuzzy comprehensive evaluation
