灰色模型在转炉终点温度预测中的应用

以邯钢邯宝炼钢厂转炉历史生产数据为基础,应用小波分析与灰色模型相结合的方法建立了转炉终点温度预测模型,在目前预测方法的基础上提出了一种新的思路。通过MATLAB中的小波变换工具,剔除数据“噪声”后分为趋势数据和波动数据,并以24个炉次为待测样本,利用灰色模型对转炉终点温度进行预测。结果表明,预测误差在[－10, 10]区间的占据95.83%,能够满足生产要求,从而验证了模型的有效性。     

两种转炉炼钢终点控制模型的比较

分析了转炉炼钢静态控制与动态控制模型的方法、作用及功能关系。针对中小型炼钢转炉检测设备受限制的现状,从建模原理、方法特点和应用效果方面对建立的基于遗传算法-神经网络混合算法(CA-BP)的转炉炼钢终点优化控制静态模型和转炉炼钢终点预测的准动态控制模型进行了比较,认为混合算法可以提高静态模型的预测效果,而准动态模型可以反映转炉冶炼的动态操作和加料过程对一次倒炉时间和终点的影响。     

转炉冶炼终点静态控制预测模型

基于天津天铁冶金集团30t转炉炼钢实际生产数据，首先建立了转炉炼钢终点静态控制的吹氧量及矿石用量统计模型，其预测100个炉次吹氧量和矿石用量平均相对误差分别为0．58％及10．4％。考虑到影响终点钢水温度和碳含量的因素比较复杂，设计了预测钢水终点温度和碳含量的人工神经网格模型，利用Levenberg-Marquardt算法和257个炉次的实际生产数据进行了模型训练，并对另外100个炉次的终点钢水温度及碳含量进行了预测，在终点钢水温度为1646-1698℃和终点碳质量分数为0．033％～0．128％的范围内，得到的终点碳温双命中率为55％。     

基于机器学习的转炉热损失率预测

转炉热损失率是影响物料消耗量预测精度的重要参数之一，利用某钢厂150 t转炉1 900炉次冶炼历史生产数据，在热损失率计算的基础上，采用机器学习算法实现了转炉热损失率的准确预测。预测结果表明，相比于支持向量回归(support vector regression, SVR)和随机森林(random forest, RF)算法，轻量级梯度提升机(light gradient boosting machine, LightGBM)算法的预测精度最高；考虑上炉次的影响，增加上炉次冶炼终点温度变量后，LightGBM算法的决定系数R²由0.89提高到0.93,在±0.005、±0.01范围内，热损失率预测命中率分别由85%、89%提高到90%、93%;另外，通过算法内部参数优化可进一步提高模型预测精度，对于LightGBM算法，决定系数R²和均方根误差E_RMS(root mean square error, RMSE)进一步分别达到了0.94、0.009,在±0.005、±0.01范围内热损失率预测命中率进一步分别提高到91%、...     

基于神经网络的转炉冶炼终点锰、磷静态预测算法

在国内重工业领域中,很多钢铁企业所采用的转炉大部分为最小型的转炉,由于容量有限无法对转炉冶炼结束时的锰、磷进行静态预测,进行影响了冶炼的精度。然而,传统算法用于实现锰和磷的冶炼终点。因此,充分利用最近开发的人工神经网络技术,基于Visual Basic编程语言,神经网络模型用于预测转炉冶炼结束时的锰和磷状态。针对半钢炼钢分开建立锰、磷含量、温度预测模型,确定输入层参数有37个,中间隐藏层参数有30个,输出层参数有两个3层BP神经网络。模型在30 000炉样本的基础上做数据训练,对权值、阈值进行修正,并保存100炉未训练过的学习样本作为模型网络训练依据,对转炉冶炼进行在线训练,通过训练的模型可以很好的适应转炉冶炼多变的生产条件。     

80 t转炉终点预报TSVR模型精度

 转炉炼钢是一个复杂的高温物理化学反应过程。在冶炼过程中不能连续检测钢的成分。所以,准确地预报终点的碳质量分数和温度对于提高终点命中率是非常有意义的。基于广西某钢厂80 t转炉炼钢实际生产数据,建立了终点碳质量分数和终点温度的孪生支持向量回归机(TSVR)预测模型,对100个炉次的实际生产数据进行了模型的训练,另外30个炉次的数据用于验证模型的精度。结果表明,预测误差Δω([C])≤ 0.01%的命中率为93.3%,Δt≤15 ℃的命中率为96.7%,双命中率为90%。与BP神经网络模型相比,TSVR模型的终点碳质量分数和终点温度命中率均比BP神经网络模型高。     

转炉冶炼终点磷的预报模型

根据现场的冶炼工艺和生产数据，研究了转炉冶炼终点磷的预报方法；采用机理模型和偏最小二乘模型分析了终点磷的影响因素；建立了基于偏最小二乘回归的转炉终点磷的预报模型。结果表明，训练样本在终点磷的绝对误差为±0.003%，在控制精度范围内命中率达87.26%，测试样本在此精度范围内的命中率达到8627%。     

涟钢LF精炼温度变化研究

采用多元回归分析方法建立了涟钢210转炉厂LF钢包炉精炼终点钢水温度的变化模型,应用该模型对LF精炼终点钢水温度进行预测,对预测结果进行了统计分析,结果表明该模型对LF钢包炉精炼终点温度的预测误差较小,能对现场产生指导意义。     

通过炉气分析实现转炉连续控制

应用红外测试仪测得的炉气数据,根据转炉熔池物料平衡和热平衡,通过模型计算,论证了炉气分析在转炉连续控制中对碳和温度预报的可行性,并得出结论;炉气分析对终点戳w(C)≤0.06%炉次的碳预报命中率较高,适用于低碳钢冶炼.但该模型进行温度预报的数据来源均为间接数据,命中率较低.     

转炉炉气分析技术在本钢的初步应用

结合本钢炉气分析在线系统,建立了应用物料平衡原理预测碳含量及根据反应平衡原理预测温度变化的转炉动态模型.模型预测结果表明:熔池中的碳含量是炉气流量及CO和CO2分压的函数,其中初始碳含量的准确程度对模型的最终预测结果影响最大.熔池温度是碳含量及炉气中CO和CO2分压的函数,其中吹炼过程的平稳性是模型最终预测结果准确获得的保证.并应用本模型对1 530炉次的历史炉气数据进行了离线分析与终点检测结果的比较,收到预期效果.     

基于BP神经网络和回归的脱硫粉剂预报模型

分析了铁水脱硫时铁水温度、铁水量、初始硫含量、脱硫后硫含量对镁粉耗量的影响,表明:随铁水温度增加镁粉耗量随之增加;随脱硫后硫含量的降低,镁粉耗量明显增加且增幅逐步扩大,为降低成本,脱硫深度应控制合理。为确定合适的粉剂用量,建立了基于BP神经网络和回归的铁水脱硫粉剂预报模型,其中BP神经网络模型是粉剂模型的主输出,回归模型用于限定输出范围。铁水脱硫粉剂预报模型已实现了在线控制,无需人工干预,达到了较好的应用效果。当偏差区间为[-0.001 5%,0.001 5%]时,脱硫后硫含量的符合率为90.85%,可有效实现脱硫后硫含量的控制。     

Nutritional and socioeconomic factors in relation to prostate cancer mortality: a cross-national study

BACKGROUND: Large international variations in rates of prostate cancer incidence and mortality suggest that environmental factors have a strong influence on the development of this disease. The purpose of this study was to identify predictive variables for prostate cancer mortality in data from 59 countries. METHODS: Data on prostate cancer mortality, food consumption, tobacco use, socioeconomic factors, reproductive factors, and health indicators were obtained from United Nations sources. Linear regression models were fit to these data. The influence of each variable fit in the regression models was assessed by multiplying the regression coefficient b by the 75th (X75) and 25th (X25) percentile values of the variable. The difference, bX75 - bX25, is the estimated effect of the variable across its interquartile range on mortality rates measured as deaths per 100000 males aged 45-74 years. Reported P values are two-sided. RESULTS: Prostate cancer mortality was inversely associated with estimated consumption of cereals (bX75 - bX25 = -7.31 deaths; P = .001), nuts and oilseeds (bX75 - bX25 = -1.72 deaths; P = .003), and fish (bX75 - bX25 = -1.47 deaths; P = .001). In the 42 countries for which we had appropriate data, soy products were found to be significantly protective (P = .0001), with an effect size per kilocalorie at least four times as large as that of any other dietary factor. Besides variables related to diet, we observed an association between prostate cancer mortality rates and a composite of other health-related, sanitation, and economic variables (P = .003). CONCLUSIONS: The specific food-related results from this study are consistent with previous information and support the current dietary guidelines and hypothesis that grains, cereals, and nuts are protective against prostate cancer. The findings also provide a rationale for future study of soy products in prostate cancer prevention trials.     

一种自适应差分进化算法在煤气分配中的应用

通过分析钢铁生产过程中主要能耗设备的煤气使用特征及煤气分配过程中的约束,建立了以煤气消耗成本、放散成本及发电收益之和最小化的数学模型,提出了一种自适应混合差分进化算法(AHDE)进行求解.算法利用蚁群算法在路径选择能力方面的优势,构造算法中差分策略选择机制提高算法的性能.通过对标准测试函数和对实际生产数据的仿真实验,验证算法的有效性.     

Optimising process parameters for gas tungsten arc welding of an austenitic stainless steel using genetic algorithm

Automated Gas Tungsten Arc Welding (GTAW) with filler wire addition using a wire feeder is a candidate process for welding of 316LN austenitic stainless steel, which is the major structural material for the Indian 500 MWe Fast Breeder Reactors. In GTAW, the quality of the weld is characterized by the weld-bead geometry as it influences the mechanical properties and its performance during service. This paper discusses the development of computational model using genetic algorithm for determining the optimum/near-optimum GTAW process parameters for obtaining the target weld-bead profile during automatic welding of 316LN stainless steel. Using the experimental data generated on the influence of process variables on weld-bead geometry, regression models correlating the weld-bead shape parameters with the process parameters were developed for determining the objective function in genetic algorithm. Close agreement was achieved between the target weld-bead profile and the model-computed weld-bead profile. This study has shown that use of genetic algorithm is an appropriate methodology for optimising process parameters to obtain target weld-bead profile in GTAW with wire feeder of 316LN stainless steel.     

Covariates of infant mortality in China: an exploratory approach

To elucidate the nature of the relationship between infant mortality in China and a variety of covariates using data from the 2/1000 Chinese Fertility Survey, we use a logistic regression model where the covariates are transformed with the help of Alternating Conditional Expectation (ACE) algorithm. This approach is used to overcome the general problem in multivariate regression analysis of coding the independent variables so that relationship between independent variables and response variables is best described, rather than coding such variables in an arbitrary way. The study demonstrates the procedures and usefulness of the ACE guided transformation in multivariate analysis. The transformed covariates are then used to estimate the effects of a series of socioeconomic and demographic factors collected in the study of infant death in China. The study shows that after appropriate transformations, all the demographic and socioeconomic variables selected have statistically significant and direct influence on infant death.     

基于支持向量机的高炉煤气利用率预测建模

煤气利用率是反映高炉能耗和平稳运行的重要指标。为了实现对高炉煤气利用率的准确预测,首先依据最大信息系数选择合适的输入参数,分别选取次于该状态参数时刻1 和2 h后的煤气利用率作为输出参数,并在建模之前对数据进行标准化处理。在此基础上建立基于支持向量回归(SVR)的高炉煤气利用率预测模型,并利用高炉的部分生产数据将该模型的预测结果与多层感知器(MLP)模型进行对比。最终预测结果表明,SVR模型在预测1和2 h后的煤气利用率时精确度更高,达到了更好的预测效果。     

线性回归模型在吨钢综合能耗分析中的应用

文章以包钢过去几年的吨钢综合能耗数据及其趋势变化为基础,利用线性回归模型对其进行吨钢综合能耗分析,给出未来3年吨钢综合能耗的理想预测值。     

基于定量关联规则树的分类及回归预测算法

为了解决基于Apriori的分类关联规则算法挖掘数值型数据时效率和准确率偏低的问题,提出基于定量关联规则树的分类及回归预测算法.采用改进的定量关联规则算法挖掘数值型数据生成关联规则库,并基于关联规则树结构实现分类及回归预测.研究结果表明:改进的Apriori定量关联规则挖掘算法提高了分类预测的准确率并降低了计算复杂度;而采用关联规则树结构可使分类与回归预测时间明显加快,提高了样本匹配学习的速度.      

Prediction of Mechanical Properties of Hot Rolled Strip by Using Semi-Parametric Single-Index Model

 A semi-parametric single-index model based approach was proposed for prediction of mechanical properties of hot rolled strip. Based on industrial production data, a semi-parametric single-index model was developed by choosing the appropriate kernel function and window width to predict the yield strength, tensile strength and elongation. When data samples are limited, compared with regression method and neural network method, the prediction results show that the semi-parametric single-index model based method is more adaptive and the prediction performance is superior to those by both regression and neural network methods.     

The effect of cigarette taxes on cigarette consumption

OBJECTIVES: This paper reexamines the work of Meier and Licari in a previous issue of the Journal. METHODS: The impact of excise taxes on cigarette consumption and sales was measured via standard regression analysis. RESULTS: The 1983 federal tax increase is shown to have an anomalous effect on the regression results. When those data are excluded, there is no significant difference between state and federal tax increases. Further investigation suggests that firms raised cigarette prices substantially in the years surrounding the 1983 federal tax increase, which accounts for the relatively large decrease in consumption during this period. CONCLUSIONS: Federal excise taxes per se do not appear to be more effective than state excise taxes in terms of reducing cigarette consumption. The reaction of cigarette firms to government policies appears to be an important determinant of the success of antismoking initiatives.