基于变量选择的转炉炼钢终点预报模型

转炉炼钢的终点预报模型对于钢水终点碳含量和温度的命中非常重要.针对高维输入不利于建立精确模型的问题,使用互信息方法对预报模型输入变量进行选择.为了区分各输入变量对输出的不同重要程度,对各输入变量进行加权处理,并采用微粒群算法对权值进行优化.最后,使用支持向量机方法建立转炉炼钢终点碳含量和温度预报模型.对一座180t转炉实际生产数据进行仿真,结果表明,合理的变量选择和加权处理能有效提高模型的预报精度.     

烧结终点预报神经网络样本优选与系统建模

烧结终点预报对于提高烧结矿强度和产量、降低能耗具有重要意义,但是烧结终点状态受多种因素影响,无法直接检测,只能由操作工依据经验进行判断,严重影响了烧结生产的稳定运行.本系统运用K均值聚类分析的样本优选方法对海量数据进行处理,选择具有代表性的样本,从而有效缩小样本空间、改善样本质量.使用风箱温度曲线计算废气温度上升点和烧结终点软测量值,以台车速度和点火温度作为输入,采用BP神经网络模型,对烧结终点位置进行预报.在实际应用中,该模型预报结果较准确地反映了烧结终点位置的变化,起到了稳定生产、节约能源的作用.     

基于BFGS算法的BP网络在烧结终点预报中的应用

烧结生产是钢铁生产过程的一个重要环节.由于烧结过程的大滞后特性,需要时烧结终点进行提前预报.基于此,建立了基于BFGS法的BP神经网络的烧结终点预报模型,并给出了详细的算法和实例.程序运行结果表明,该模型能够对烧结终点进行较为准确地预报.     

基于鲁棒相关向量机的转炉炼钢终点预报模型

针对传统相关向量机在训练过程中易受异常点影响的问题,提出了一种鲁棒相关向量机模型,并将其应用于转炉炼钢终点碳含量和温度的预报.通过为每一个训练样本设定独立的噪声方差系数,并使其在训练过程中随模型预测误差的增大而逐渐减小来降低异常点的影响,同时依据贝叶斯证据框架给出了模型超参数的迭代计算公式,进行参数的优化.使用标准测试数据和转炉炼钢实际生产数据进行仿真,结果表明本文模型具有较好的预报精度和鲁棒性.     

基于混合递阶遗传神经网络的转炉温度预报

针对转炉控制中对吹炼终点温度的控制问题,提出了基于混合递阶遗传RBF神经网络(HGA-RBF)的转炉炼钢终点温度预报模型.研究了RBF网络的特点,用递阶遗传算法克服了网络的结构和参数选择的随机性问题;并结合最小二乘法,提高了收敛速度.仿真结果表明,此算法在一定程度上提高了RBF网络的优化收敛速度和训练测试精度.某钢铁公司提供的实际冶炼数据试验,也证明了该模型预报精度较高,对提高生产的质量有重要意义.     

电弧炉终点温度的GASVR_GM预报模型研究

电弧炉终点温度是炼钢过程中的重要指标之一,决定了钢水的质量和整体成本.电弧炉终点温度预报模型的建立是实现炼钢自动化的重要环节,为了得到高精度的终点温度预报值,提出了一种GASVR_GM的钢水终点温度预报模型.该模型以定量因素为主,采用遗传算法优化的支持向量回归机预报终点温度;再运用灰色模型进行预报误差的补偿,解决非定量因素的影响,实现滚动预报.试验仿真表明,与智能软测量方法相比,GASVR_GM预报模型具有更高的精度和鲁棒性.     

基于混合递阶遗传神经网络的转炉温度预报

针对转炉控制中对吹炼终点温度的控制问题，提出了基于混合递阶遗传RBF神经网络（HGA-RBF）的转炉炼钢终点温度预报模型。研究了RBF网络的特点，用递阶遗传算法克服了网络的结构和参数选择的随机性问题；并结合最小二乘法，提高了收敛速度。仿真结果表明，此算法在一定程度上提高了RBF网络的优化收敛速度和训练测试精度。某钢铁公司提供的实际冶炼数据试验。也证明了该模型预报精度较高，对提高生产的质量有重要意义。     

基于多任务学习的炼钢终点预测方法

钢水质量通常根据终点命中率来判断,但炼钢过程影响因素众多,机理分析难以准确预测终点温度和含碳量,鉴于此,提出一种由数据驱动的多任务学习（MTL）炼钢终点预测方法。首先,分析并提取炼钢过程的输入和输出要素,结合炼钢两阶段吹炼特点选择多个子学习任务;其次,根据子任务与终点参数的相关性选择合适的子任务,提升终点预测的准确度并构建多任务学习模型,再对模型输出结果进行二次优化;最后,通过近端梯度算法对处理后的生产数据进行模型训练,获取多任务学习模型的过程参数。以某钢厂为案例,该方法相比神经网络在终点温度12℃误差范围内和终点含碳量0.01%误差内的准确度提升了10%,误差范围6℃和0.005%的预测准确度分别提升了11%和7%。实验结果表明,多任务学习在实际中能够提升终点预测的准确性。     

一种基于增量式神经网络的转炉静态控制模型

针对中小型转炉不宜增设副枪、难以对钢水成分和温度进行连续检测、难以建立动态模型的实际情况,本文将传统增量模型和神经网络模型有机结合,提出了一种基于增量式神经网络的转炉静态控制模型,对钢水终点进行控制。在该模型引入了RBF神经网络对钢水终点温度和碳含量进行实时预报,使得对增量式神经网络控制模型的训练以预报模型的输出值与所要求的钢水终点温度和碳含量之差为最小,克服了常规静态控制模型存在的不足,改善了控制效果,提高了炼钢一倒命中率。     

炼铜转炉吹炼终点的神经网络和自适应残差补偿组合预报模型

提出了基于改进的BP神经网络学习算法和自适应残差补偿算法的炼铜转炉吹炼终点组合预报模型. 利用某厂实际生产数据进行仿真运行的结果表明, 本文建立的模型具有较高的预报精度和较强的实用性, 可用于指导生产实践.     

Gray-box modeling for prediction and control of molten steel temperature in tundish

To realize stable production in the steel industry, it is important to control molten steel temperature in a continuous casting process. The present work aims to provide a general framework of gray-box modeling and to develop a gray-box model that predicts and controls molten steel temperature in a tundish (TD temp) with high accuracy. Since the adopted first-principle model (physical model) cannot accurately describe uncertainties such as degradation of ladles, their overall heat transfer coefficient, which is a parameter in the first-principle model, is optimized for each past batch separately, then the parameter is modeled as a function of process variables through a statistical modeling method, random forests. Such a model is termed as a serial gray-box model. Prediction errors of the first-principle model or the serial gray-box model can be compensated by using another statistical model; this approach derives a parallel gray-box model or a combined gray-box model. In addition, the developed gray-box models are used to determine the optimal molten steel temperature in the Ruhrstahl–Heraeus degassing process from the target TD temp, since the continuous casting process has no manipulated variable to directly control TD temp. The proposed modeling and control strategy is validated through its application to real operation data at a steel work. The results show that the combined gray-box model achieves the best performance in prediction and control of TD temp and satisfies the requirement for its industrial application.     

基于RBF神经网络的钢坯温度预报软测量模型研究

钢坯加热过程是钢铁企业热轧生产中非常重要的工艺环节。钢坯温度预报模型是实现加热炉优化控制的重要基础,用常规仪器很难直接测量出钢坯温度。给出了基于RBF神经网络的软测量模型结构,对钢坯温度进行预报的仿真结果。     

Neural network based model predictive control for a steel pickling process

A multi-layer feedforward neural network model based predictive control scheme is developed for a multivariable nonlinear steel pickling process in this paper. In the acid baths three variables under controlled are the hydrochloric acid concentrations. The baths exhibit the normal features of an industrial system such as nonlinear dynamics and multi-effects among variables. In the modeling, multiple input, single-output recurrent neural network subsystem models are developed using input–output data sets obtaining from mathematical model simulation. The Levenberg–Marquardt algorithm is used to train the process models. In the control (MPC) algorithm, the feedforward neural network models are used to predict the state variables over a prediction horizon within the model predictive control algorithm for searching the optimal control actions via sequential quadratic programming. The proposed algorithm is tested for control of a steel pickling process in several cases in simulation such as for set point tracking, disturbance, model mismatch and presence of noise. The results for the neural network model predictive control (NNMPC) overall show better performance in the control of the system over the conventional PI controller in all cases.     

真空感应炉终点碳含量预报

针对冶炼过程中碳含量不能直接测定的不足。采用RBF神经网络对真空感应炉的终点碳含量进行预报，在第一次预报时，初步计算出冶炼到达终点的时间和终点的碳含量；经过二次预报进行误差校正，使结果更加精确，结合现场120组数据进行学习和预报，预报命中率较高．实验结果表明，采用该方法预报碳含量可以取得良好的效果。     

Endpoint prediction model for basic oxygen furnace steel-making based on membrane algorithm evolving extreme learning machine

The endpoint parameters of molten steel, such as the steel temperature and the carbon content, directly affect the quality of the production steel. Moreover, these endpoint results cannot be the online continuous measurement in time. To solve the above-mentioned problems, an anti-jamming endpoint prediction model is proposed to predict the endpoint parameters of molten steel. More specifically, the model is constructed on the parameters of extreme learning machine (ELM) adaptively adjusted by the evolutionary membrane algorithm with the global optimization ability. In other words, the evolutionary membrane algorithm may find the suitable parameters of an ELM model which reduces the incidence of the overfitting of ELM affected by the noise in the actual data. Finally, the proposed model is applied to predict the endpoint parameters of molten steel in steel-making. In the simulation experiments, two test problems, including ‘SinC’ function with the Gaussian noise and the actual production data of basic oxygen furnace (BOF) steel-making, are employed to evaluate the performance of the proposed model. The results indicate that the proposed model has good prediction accuracy and robustness in the data with noise. Therefore, the proposed model has good application prospects in the industrial field.     

支持向量机在高炉铁水温度预测中的应用

铁水温度是高炉冶炼过程的关键参数,是影响高炉稳定顺行及节能降耗的重要指标。以高炉炉内热状态的重要指示剂-铁水温度为研究对象,在综合利用K-means 聚类和支持向量机方法的各自优势和互补情况下,提出一种基于K-means 聚类的支持向量机预测铁水温度的方法,该方法首先将训练样本数据分为m 类,建立m 个支持向量机回归预测模型,同时采用粒子群算法优化模型参数; 其次建立m 个判别函数,判别待预测样本数据属于哪一类;最后将待预测样本数据代入相应类的回归模型中进行预测。相比标准支持向量机预测,得到了较高的预测精度。     

Dynamic control model of BOF steelmaking process based on ANFIS and robust relevance vector machine

This study concerns with the control of basic oxygen furnace (BOF) steelmaking process and proposes a dynamic control model based on adaptive-network-based fuzzy inference system (ANFIS) and robust relevance vector machine (RRVM). The model aims to control the second blow period of BOF steelmaking and consists of two parts, the first of which is to calculate the values of control variables, viz., the amounts of oxygen and coolant requirement, and the other is to predict the endpoint carbon content and temperature of molten steel. In the first part, an ANFIS classifier is primarily constructed to determine whether coolant should be added or not, then an ANFIS regression model is utilized to calculate the amounts of oxygen and coolant. In the second part, a novel robust relevance vector machine is presented to predict the endpoint. RRVM solves the problem of sensitivity to outlier characteristic of classical relevance vector machine, thus obtaining higher prediction accuracy. The key idea of the proposed RRVM is to introduce individual noise variance coefficient to each training sample. In the process of training, the noise variance coefficients of outliers gradually decrease so as to reduce the impact of outliers and improve the robustness of the model. Simulations on industrial data show that the proposed dynamic control model yields good results on the oxygen and coolant calculation as well as endpoint prediction. It is promising to be utilized in practical BOF steelmaking process.     

基于PLC的铁水脱硫专家模型的设计

铁水脱硫是高炉炼钢中必要工艺,用以降低铁水中杂质硫的含量.设计了基于PLC的铁水脱硫专家模型系统,并引入BP人工神经网络,实现了铁水脱硫过程的自动控制,降低能源消耗和环境污染,提高了生产效率和钢材质量.