基于炉口辐射光谱支持向量机回归的转炉终点碳含量检测

转炉终点碳含量在线检测对实现炼钢终点准确控制,提高钢铁产品质量,降低能耗,减少废气排放具有重要意义。针对转炉冶炼终点控制及碳含量检测的难题,研究了一种新的基于炉口辐射光谱分析的非接触式在线碳含量检测方法。方法基于辐射光谱的支持向量机回归(SVR)实现转炉终点前过程的碳含量预测。通过远距离光谱采集系统获取炉口火焰光谱信息,基于冶炼过程炉口火焰辐射光谱变化规律的分析,分别提取了表征辐射光谱整体特征的两个参数即总谱宽和辐射峰值、以及表征发射谱的三个特征波长600,630和775 nm处的幅值作为支持向量机的输入,结合脱碳理论和实测碳值拟合重构的脱碳函数曲线作为支持向量机的输出,利用支持向量机回归方法建立光谱分布与碳含量的关系模型。通过训练样本集和测试集循环优化确定模型最佳参数。设计的仪器和优选的模型已安装在转炉生产现场长时间运行,现场实验结果表明,终点碳含量检测准确率为90.2%,测量时间小于0.3 s,可实时在线检测,能够满足生产需求,为转炉冶炼终点的精确控制提供了重要依据。     

WCARS-ISPA火焰光谱特征选择的转炉炼钢终点预测

对转炉炼钢终点的实时精准控制能够有效提高钢铁产出的质量,炉口火焰光谱在炼钢不同时期的变化明显,对其进行分析处理并与机器学习方法相结合可有效用于炼钢终点的实时控制。针对炉口火焰光谱数据量大、现有方法对光谱特征提取在可信度和实时性上不足的缺陷,提出一种基于窗口竞争性自适应重加权采样（WCARS）结合迭代式连续投影算法（ISPA）的光谱特征波长选择方法,该方法在有效解决模型过拟合问题的同时,能够降低高维数据计算的复杂度。将火焰光谱数据沿波长方向进行窗口划分后,使用CARS进行计算选出特征窗口波段,再将迭代式选择与传统连续投影算法相结合,通过重复迭代精选出特征波长,在此基础上使用支持向量机回归（SVR）建立炼钢终点碳含量预测模型。实验采集363组炼钢后期的炉口火焰光谱数据作为样本,并对其进行Savitzky-Golay平滑预处理。使用WCARS-ISPA算法从全光谱数据中选出10个特征波长作为SVR模型的输入,碳含量为模型输出,Kennard-stone算法对训练集和测试集进行划分,选择碳含量的平均预测误差、预测误差在±2%以内的命中率以及运行30次的平均时间作为模型评价指标。实验结果显示,模型的平均碳含量预测误差为1.413 2%,命中率高达90.63%,运行时间为0.019 679 s。与使用全光谱和WCARS-ISPA,CARS-SPA,WCARS和SPA四种不同特征选择方法选出的特征波长建模得到的结果进行对比,基于WCARS-ISPA方法选出的特征波长建立的终点碳含量预测模型误差最小、命中率最高。提出一种新的炉口火焰光谱特征波长提取方法,使用窗口竞争性自适应重加权采样结合迭代式连续投影算法选取特征波长,并在此基础上建立转炉炼钢终点碳含量预测模型,实验结果表明,该方法能够有效提取火焰光谱特征,所建模型能够对转炉炼钢终点进行准确预测,满足工业生产的实时控制要求,为实际生产提供可靠帮助。     

A model of basic oxygen furnace (BOF) end-point prediction based on spectrum information of the furnace flame with support vector machine (SVM)

In the common basic oxygen furnace technique, we often get the information of furnace flame by our eyes, from which we could estimate the content of the carbon in the molten iron. In order to using an optical device to estimate the end point of BOF accurately, we design an arithmetic to predict the end point. The variety of the furnace flame essentially is the variety of the spectrum. So we proceed with the variety of spectrum distributing, distilling some parameters which could depict the variety of spectrum, combining with the support vector machine (SVM) to compose a system which could predict the end point of BOF relatively accuracy. The result of the experiment shows that the model meets the requirements of end-point judgment online. Furthermore, by discussing with the different output function of SVM, we can confirm that we can measure the content of carbon in the molten iron in real time by using this model.     

Basic oxygen furnace steelmaking end-point prediction based on computer vision and general regression neural network

End-point prediction is one of the most difficult problems in basic oxygen furnace (BOF) steelmaking process. To address this problem, some researchers have proposed some methods based on flame image processing and pattern classification. Because of the dynamically changing flame and real-time needs during the blowing process, there are still some issues that need to be solved. We propose a novel method based on accurate and fast multi flame features extraction and general regression neural network (GRNN). Firstly, flame images were acquired, and then the background of each image was removed via color similarity determination algorithm; secondly, color, texture, and boundary features were extracted; the fast and robust boundary and texture features were extracted by using the proposed methods, and these features were tested for their validity to the end-point prediction via comparing them with some other similar methods; finally, the prediction model was built using multi-features and GRNN. The experimental results demonstrated that it is accurate and fast to use the proposed method to the BOF end-point predict.     

基于炉口火焰辐射测温的转炉终温预测研究

转炉炼钢的铁水终点温度是衡量铁水是否符合出钢要求的一个重要指标,目前仍主要依靠人工经验进行控制,其预测精度难以保证。为了能够有效的预测终点铁水的温度,并能满足不同的转炉炉口尺寸的要求,提出了一种基于炉口火焰温度测量的转炉终点铁水温度预测的新方法。利用可以采集可见光和近红外波段的微型光谱仪作为关键器件构建了炉口火焰辐射信息的采集系统,实现对转炉炼钢现场的炉口火焰辐射信息进行实时、有效的采集。同时综合考虑了炉口火焰的实际温度范围和现场的测量距离,通过改进的系统标定方法,对所构建的辐射信息采集系统进行标定,并利用双波长法获得了炉口火焰的辐射温度。鉴于炉口的火焰温度与炉内的铁水温度之间存在着一定的联系,还利用所测得的火焰温度以及与铁水温度相关的转炉吹炼过程参数,建立了基于支持向量机的炉内铁水温度的回归预测模型。对68炉现场采集到的转炉吹炼数据进行验证实验后表明该方法的预测精度要优于人工经验,且接近采用副枪的铁水温度测量方式,可为中小型转炉的终点温度预测提供一种有效可行的方案。     

基于多类支持向量机的恒星光谱分类方法

支持向量机作为一种经典的分类方法被广泛应用于恒星光谱分类领域。该方法在实际应用中取得了较为理想的分类效果,但其面临无法解决多分类问题的挑战。在支持向量机的基础上,提出多类支持向量机,建立基于多类支持向量机的恒星光谱分类模型。该方法的最大优势是经过一次分类过程,可以确定多类样本的类属。SDSS DR8恒星光谱数据上的比较实验表明,本研究所提的方法较之已有多分类方法在分类性能上有一定的提升。     

Intelligent Fault Identification for Rolling Bearings Fusing Average Refined Composite Multiscale Dispersion Entropy-Assisted Feature Extraction and SVM with Multi-Strategy Enhanced Swarm Optimization

Rolling bearings act as key parts in many items of mechanical equipment and any abnormality will affect the normal operation of the entire apparatus. To diagnose the faults of rolling bearings effectively, a novel fault identification method is proposed by merging variational mode decomposition (VMD), average refined composite multiscale dispersion entropy (ARCMDE) and support vector machine (SVM) optimized by multistrategy enhanced swarm optimization in this paper. Firstly, the vibration signals are decomposed into different series of intrinsic mode functions (IMFs) based on VMD with the center frequency observation method. Subsequently, the proposed ARCMDE, fusing the superiorities of DE and average refined composite multiscale procedure, is employed to enhance the ability of the multiscale fault-feature extraction from the IMFs. Afterwards, grey wolf optimization (GWO), enhanced by multistrategy including levy flight, cosine factor and polynomial mutation strategies (LCPGWO), is proposed to optimize the penalty factor C and kernel parameter g of SVM. Then, the optimized SVM model is trained to identify the fault type of samples based on features extracted by ARCMDE. Finally, the application experiment and contrastive analysis verify the effectiveness of the proposed VMD-ARCMDE-LCPGWO-SVM method.     

Variable order spherical harmonic expansion scheme for the radiative transport equation using finite elements

We propose the P_N approximation based on a finite element framework for solving the radiative transport equation with optical tomography as the primary application area. The key idea is to employ a variable order spherical harmonic expansion for angular discretization based on the proximity to the source and the local scattering coefficient. The proposed scheme is shown to be computationally efficient compared to employing homogeneously high orders of expansion everywhere in the domain. In addition the numerical method is shown to accurately describe the void regions encountered in the forward modeling of real-life specimens such as infant brains. The accuracy of the method is demonstrated over three model problems where the P_N approximation is compared against Monte Carlo simulations and other state-of-the-art methods.     

Detailed and simplified kinetic models of n-dodecane oxidation: The role of fuel cracking in aliphatic hydrocarbon combustion

A detailed kinetic model is proposed for the combustion of normal alkanes up to n-dodecane above 850 K. The model was validated against experimental data, including fuel pyrolysis in plug flow and jet-stirred reactors, laminar flame speeds, and ignition delay times behind reflected shock waves, with n-dodecane being the emphasis. Analysis of the computational results reveal that for a wide range of combustion conditions, the kinetics of fuel cracking to form smaller molecular fragments is fast and may be decoupled from the oxidation kinetics of the fragments. Subsequently, a simplified model containing a minimal set of 4 species and 20 reaction steps was developed to predict the fuel pyrolysis rate and product distribution. Combined with the base C₁-C₄ model, the simplified model predicts fuel pyrolysis rate and product distribution, laminar flame speeds, and ignition delays as close as the detailed reaction model.     

基于炉口辐射和改进神经网络的转炉终点预测模型

针对国内外转炉炼钢终点控制的现状,建立了一种用于终点预测的神经网络模型.以炉口辐射信息获取系统为实验平台,运用光纤谱分复用和颜色空间模型转换技术.分析发现了光谱与图像信息特征量在吹炼过程中晕现出中前期类似、末期相反的规律.从得到的特征规律曲线中选用一些关键特征量,在改进的修正系数算法基础上,进行了模型的训练和预测分析.实验结果表明:响应时间在2 s以内,满足快速判定的时间要求;改进算法的模型预测精度高于常规算法,该系统可以止常工作在转炉炼钢的恶劣环境下,达到了预期效果.     

The effects of the choice of the k-interval number on radiative calculations

The selection of the number of k-interval is a foundation to correlated k-distribution method and the problem of how to do it still remains unsettled. It is pointed out by numerical computation in this work that choosing the number of k-interval is a major factor affecting accuracy and speed in radiative calculation. To increase the number of k-interval is an efficient method to improve the accuracy. However, it is found by this study that there exists a saturation of the accuracy to an increase of the number. The optimal rules on the number of k-interval choosing are proposed in the paper. Then, five versions on atmospheric absorption by gases appropriate for GCMs are given according to them.     

Skeletal mechanism generation with CSP and validation for premixed n-heptane flames

An automated procedure has been previously developed to generate simplified skeletal reaction mechanisms for the combustion of n-heptane/air mixtures at equivalence ratios between 0.5 and 2.0 and different pressures. The algorithm is based on a Computational Singular Perturbation (CSP)-generated database of importance indices computed from homogeneous n-heptane/air ignition solutions. In this paper, we examine the accuracy of these simplified mechanisms when they are used for modeling laminar n-heptane/air premixed flames. The objective is to evaluate the accuracy of the simplified models when transport processes lead to local mixture compositions that are not necessarily part of the comprehensive homogeneous ignition databases. The detailed mechanism was developed by Curran et al. and involves 560 species and 2538 reactions. The smallest skeletal mechanism considered consists of 66 species and 326 reactions. We show that these skeletal mechanisms yield good agreement with the detailed model for premixed n-heptane flames, over a wide range of equivalence ratios and pressures, for global flame properties. They also exhibit good accuracy in predicting certain elements of internal flame structure, especially the profiles of temperature and major chemical species. On the other hand, we find larger errors in the concentrations of many minor/radical species, particularly in the region where low-temperature chemistry plays a significant role. We also observe that the low-temperature chemistry of n-heptane can play an important role at very lean or very rich mixtures, reaching these limits first at high pressure. This has implications to numerical simulations of non-premixed flames where these lean and rich regions occur naturally.     

基于可见光谱和支持向量机的黄瓜叶部病害识别方法研究

以黄瓜叶部病害作为研究对象,基于可见光谱反射率差异识别黄瓜叶部病害,研究基于SVM的黄瓜叶部病害识别预测模型。采用小波变换进行数据预处理;选取Otsu、边缘分割法和K均值聚类三类分割方法进行病斑分割,比较错分率和运行时间,K均值聚类方法更适合黄瓜叶部病斑分割;提取纹理、颜色和形状特征参数,共15个特征参数;通过交叉验证选择最优参数c和g,对核函数参数进行优化处理,并通过比较线性核、多项式核、RBF核等不同核函数情况下SVM的正确识别率,确定RBF核SVM模式识别方法能够更精准地识别黄瓜叶部病害。并将基于SVM与另外两种常见的黄瓜叶部病害识别方法,BP神经网络和模糊聚类进行比较,结果表明,基于SVM的识别模型对霜霉病的正确识别率为95%,白粉病和褐斑病的正确识别率均为90%,平均诊断正确率为92%;该模式识别方法识别效果最佳,运行时间最短,为基于可见光谱的黄瓜病害识别模型提供参考。     

ISG-FHEV等效燃油消耗最小控制策略

针对室内复杂环境下火灾识别准确率会降低的问题,提出了一种改进的粒子群算法优化支持向量机参数进行火灾火焰识别的方法。首先在 颜色空间进行火焰图像分割,对获得的火焰图像进行预处理并提取相关特征量;其次采用PSO算法搜索SVM的最优核参数和惩罚因子,并在PSO算法中加入变异操作和非线性动态调整惯性权值的方法,加快了搜索SVM最优参数的精度和速度;然后将提取的火焰各个特征量作为训练样本输入SVM模型进行训练,并建立参数优化后的SVM分类器模型;最后将待测试样本输入SVM模型进行分类识别。算法的火灾识别准确率达到94.09%,分类效果明显优于其他分类算法。仿真结果表明,改进的PSO优化SVM算法提高了火焰识别的准确率和实时性,算法的自适应性更强,误判率更低。     

A false color image fusion method based on multi-resolution color transfer in normalization YCBCR space

In this paper, a false color image fusion method for merging visible and infrared images is proposed. Firstly, the source images and reference image are decomposed respectively by Laplacian pyramid transform. Then the grayscale fused image and the difference images between the normalized source images are assigned to construct YC_BC_R components. In the color transfer step, all the three channels of the color space in each decomposition level are processed with the statistic color mapping according to the color characteristics of the corresponding sub-images of the reference image. Color transfer is designed based on the multi-resolution scheme in order to significantly improve the detailed information of the final image, and to reduce excessive saturation phenomenon to have a comparatively natural color appearance compared with the classical global-coloring algorithm. Moreover, the differencing operation between the normalized source images not only provides inter-sensor contrast to make popping the potential targets but also weakens the influence of the ambient illumination variety to a certain degree. Finally, the fused results and several metrics for evaluation of fused images subjectively and objectively illustrate that the proposed color image fusion algorithm can yield a more complete mental representation of the perceived scene, resulting in better situational awareness.     

Fault Diagnosis of a Wind Turbine Gearbox Based on Improved Variational Mode Algorithm and Information Entropy

The working environment of wind turbine gearboxes is complex, complicating the effective monitoring of their running state. In this paper, a new gearbox fault diagnosis method based on improved variational mode decomposition (IVMD), combined with time-shift multi-scale sample entropy (TSMSE) and a sparrow search algorithm-based support vector machine (SSA-SVM), is proposed. Firstly, a novel algorithm, IVMD, is presented for solving the problem where VMD parameters (K and α) need to be selected in advance, which mainly contains two steps: the maximum kurtosis index is employed to preliminarily determine a series of local optimal decomposition parameters (K and α), then from the local parameters, the global optimum parameters are selected based on the minimum energy loss coefficient (ELC). After decomposition by IVMD, the raw signal is divided into K intrinsic mode functions (IMFs), the optimal IMF(s) with abundant fault information is (are) chosen based on the minimum envelopment entropy criterion. Secondly, the time-shift technique is introduced to information entropy, the time-shift multi-scale sample entropy algorithm is applied for the analysis of the complexity of the chosen optimal IMF and extract fault feature vectors. Finally, the sparrow search algorithm, which takes the classification error rate of SVM as the fitness function, is used to adaptively optimize the SVM parameters. Next, the extracted TSMSEs are input into the SSA-SVM model as the feature vector to identify the gear signal types under different conditions. The simulation and experimental results confirm that the proposed method is feasible and superior in gearbox fault diagnosis when compared with other methods.     

A method of segment protection based on p-cycle for dynamic multicast with limited-range wavelength converter in WDM network

In the optical multicast network, node and link failures have very important influence on the network survivability which may lead to multiple destinations cannot receive data. Based on the wavelength layered-graph method, a method of efficiency-score based on heuristic algorithm of pre-configured cycle (p-Cycle) based-segment protection (ESHS) for dynamic multicast with limited-range wavelength conversion was presented in this paper. By finding the multicast tree segment protection, the total available p-Cycles are constructed for the multicast. Then we calculate each p-Cycle efficiency-score, the highest efficient-score p-Cycle is selected as the multicast route protection p-Cycle. The simulation results show that the ESHS can get higher performance than the existing ESHT algorithm, in terms of multicast request blocking probability and wavelength utilization.     

Study on the influencing factors of analyzing gold-loaded carbon by flame atomic absorption spectrometry

In this paper, the influencing factors on rapid flame atomic absorption spectrometry analysis of gold-loaded carbon, granularity and moisture content are discussed. The main errors came from three aspects. One was the inhomogeneity of gold distribution in activated carbon. The other was moisture content. The last one was the error from the determining process by flame atomic absorption spectrometry. The relationship between granularity, moisture content and the inhomogeneity of gold content was investigated by analysing real gold-loaded carbon samples. The results shows that the nonuniformity of distribution of moisture content and gold content could be eliminated if the granularity of gold-loaded carbon is higher than 74 microns. After the operating parameters of atomic absorption spectrometry were optimized, the error in the process of determining gold by FAAS was reduced. The results sampled by the proposed method are consistent with the iodimetric analysis. The accuracy of this method is investigated by adding gold standard as well. The recoveries are in the range of 91-97%. Our results indicate the improvement of the gold-loaded carbon analysis.     

Calculation of on-state I–V characteristics of LDMOSFETs based on an accurate LDD resistance modeling

In this paper, we present an I–V model for LDMOSFETs. It is based on modeling the Lightly-Doped Drain (LDD) region of the device as voltage-controlled resistors where velocity saturation effect is also taken into account. Using the LDD region model along with a model for the channel region of the device, the on-state I–V characteristic of the transistor is accurately calculated. The models for the LDD region resistors can be incorporated into a circuit simulator such as HSPICE which has an accurate model for the channel region of the transistor. The accuracy of the models is verified by comparing its results with those of a device simulator. The results show a maximum error of 1% for a wide range of voltages and overlapped LDD region lengths.     

An efficient model for laser surface hardening of hypo-eutectoid steels

This paper presents a model able to predict the austenization of hypo-eutectoid steels during very fast heat cycle such as laser hardening. Laser surface hardening is a process highly suitable for hypo-eutectoid carbon steels with carbon content below 0.6% or for low alloy steels where the critical cooling rate is reached by means of the thermal inertia of the bulk. As proposed by many authors, the severe heat cycle occurring in laser hardening leads to the pearlite to austenite microstructures transformation happening to a temperature much higher than the eutectoid temperature A_c1 and, afterwards, all the austenite predicted during the heating phase become martensite during quenching. Anyway, all these models usually generate a predicted hardness profile into the material depth with an on-off behavior or very complicated and time consumed software simulators. In this paper, a new austenization model for fast heating processes based on the austenite transformation time parameter I_p→a is proposed. By means of the I_p→a parameter it is possible to predict the typical hardness transition from the treated surface to the base material. At the same time, this new austenization model also reduces the calculation time. I_p→a was determined by experimental tests and it was postulated to be constant for low-medium carbon steels. Several experimental examples are proposed to validate the assumptions and to show the accuracy of the model.