Neural network based on adaptive chaotic gradient descending optimization algorithm and its application in matte converting process

An adaptive chaotic gradient descending optimization algorithm for single objective optimization was presented. A local minimum judged by two rules was obtained by an improved mutative-step gradient descending method. A new optimal minimum was obtained to replace the local minimum by mutative-scale chaotic search algo-rithm whose scales are magnified gradually from a small scale in order to escape local minima. The global optimal value was attained by repeatedly iterating. At last, a BP (back-propagation) neural network model for forecasting slag output in matte converting was established. The algorithm was used to train the weights of the BP neural net-work model. The simulation results with a training data set of 400 samples show that the training process can be fin-ished within 300 steps to obtain the global optimal value, and escape local minima effectively. An optimization sys-tem for operation parameters, which includes the forecasting model, is achieved, in which the output of converter increases by 6.0%, and the amount of the treated cool materials rises by 7.8% in the matte converting process.     

On-line forecasting model for zinc output based on self-tuning support vector regression and its application

An on-line forecasting model based on self-tuning support vectors regression for zinc output was put forward to maximize zinc output by adjusting operational parameters in the process of imperial smelting furnace. In this model, the mathematical model of support vector regression was converted into the same format as support vector machine for classification. Then a simplified sequential minimal optimization for classification was applied to train the regression coefficient vector α-α^* and threshold b. Sequentially penalty parameter C was tuned dynamically through forecasting result during the training process. Finally, an on-line forecasting algorithm for zinc output was proposed. The simulation result shows that in spite of a relatively small industrial data set, the effective error is less than 10% with a remarkable performance of real time. The model was applied to the optimization operation and fault diagnosis system for imperial smelting furnace.     

矩形六电极炼镍矿热熔炼炉电场的模型试验

以水力模型实验为手段对１６５００ｋＶＡ六电极炼镍矿热电炉进行了模拟试验，探讨了熔池中的电位分布、电流分布、功率分布特点及溶液浓度、熔池渣层深度、加料方式、电极插入深度、电极尺寸等因素对熔池电场分布的影响，提出了有关参数的优化值。     

PS转炉铜锍吹炼过程智能决策与终点预报系统

利用建立的基于神经网络与自适应残差补偿的组合预报模型对吹炼终点范围进行预报，采用物料衡算与模糊推理方法在线推算预报结果的校验范围。在此基础上开发炼铜转炉优化操作智能决策支持系统。应用表明，终点范围预报准确率在85％以上。转炉产量提高6％，冷料处理量提高7．8％。     

有色冶金过程的数据挖掘

介绍数据挖掘技术，分析有色冶金过程数据的特点，探讨数据挖掘技术在有色冶金过程的应用途径、重要原则及关键问题，并提出有色冶金过程数据挖掘的一般步骤。     

基于彩色CCD的比色测温校正方法

基于彩色CCD的比色测温技术是当前高温测量领域中的研究热点之一.被测物体的发射率随辐射波长变化以及CCD光谱响应特性非理想是该方法的主要误差来源.通过对比色测温原理和CCD成像原理的分析,提出了含有发射率校正系数和CCD响应带宽校正系数的比色测温公式,以减少CCD光谱响应特性非理想和被测辐射体发射率变化带来的测量误差.提出利用黑体炉实验标定CCD响应带宽校正系数a、b,利用现场实验标定发射率校正系数c.实验表明,本方法能有效地减小测温误差,具有较强的实用性.     

链篦机球团预热系统能耗分析及优化

链篦机作为“链篦机-回转窑”球团生产系统的关键设备,其运行状况直接影响生产系统的能耗、产品产量及质量.针对国内某厂链篦机球团预热系统,分别建立了各工艺段的热分析和(火用)分析模型,并根据实际生产中的热工测试数据,对其用能状况进行了计算分析.结果表明:该链篦机预热系统的整体热效率为51％～55％,(火用)效率为42％～46％,各工艺段的热效率均低于66％,(火用)效率均低于53％,废气排放能量损失是链篦机系统的主要能量损失.在此基础上,提出了链篦机球团预热系统的优化用能方法,即将抽风干燥段和预热—段排出的热风引入鼓风干燥段,用于球团干燥,可将系统的整体热效率提高到60％,(火用)效率提高到49.6％,从而有效地优化用能过程,提高用能水平,减少能源消耗.     

Fast generation method of fuzzy rules and its application to flux optimization in process of matter converting

1 INTRODUCTIONIn general ,the physical attributes of tempera-ture ,pressure and position are continuous in theprocess of industrial production. Data with contin-uous attributes are also called numerical data orquantitative data[1 3]. Numerical data can reflectthe real world,but it is very indigestible .It is al-most i mpossible to build mathematical model to op-ti mize complexindustrial process .So it is difficultto obtain steady and opti mal productionindexes ofthe process of nonferrous …     

一种基于小波分析的异常数据样本检测与修复方法

准确的数据预处理是实现正确有效的数据挖掘和基于数据样本进行系统建模的前提和关键环节．数据预处理的一项重要任务是从大量数据样本中剔除异常样本或对受污染的样本进行清洗修复，但在数据集中各数据项间关系未知的情况下。检测异常样本比较困难，为此，文章提出了一种基于小波分析的异常数据样本检测与修复方法．该方法充分利用了小波分析的多尺度、多分辨特性，能有效地实现异常数据样本的准确检测和修复．为了实现离散序列小波变换的快速计算，文章还提出了一种基于Newton-Cores公式的修正的数值积分算法．仿真结果表明，文章提出的方法切实可行，效果盘好，有很强的实用性．