矿物浮选吸附平衡模型构建与应用：精准解析矿物表面离子/药剂特征吸附

矿物浮选吸附平衡模型是准确描述矿物浮选过程中所涉及的矿物表面离子/药剂吸附平衡的数学模型，首次实现了矿物表面位点与浮选药剂吸附作用亲和性的量化解析。传统浮选理论对矿物表面作用位点缺乏科学描述，难以阐明浮选药剂的吸附作用能力及平衡状态，新浮选药剂体系开发多以“试错法”“复配法”等经验性方法为主。药剂选择性吸附作为浮选作业的核心机制，若不能准确预测其在各矿物表面的吸附行为，将严重限制浮选工艺智能化控制的发展。本文详细论述了矿物浮选吸附平衡模型的构建原理，以赤铁矿-石英和一水硬铝石-高岭石两个浮选体系为例，通过对少量实验数据的拟合计算，解析出矿物表面活性位点密度Ns、加/去质子反应常数Kt1/Kt2、药剂吸附常数Kf等矿物浮选特征常数。将解析的矿物浮选特征常数代入到平衡模型中，预测出各条件下矿物表面电性、离子及浮选药剂的吸附量，通过零电点、实验及测试等方法验证了可靠性，形成了矿物浮选过程预测系统的内核算法。该算法对各条件下每种矿物药剂吸附的初步准确预测，可在一定程度上反映其可浮性趋势，有助于缩短浮选工艺开发周期，对矿物表面药剂吸附机理研究、浮选药剂分子设计、浮选工艺流程优化及智能控制等具有重...     

基于舍一交叉验证优化最小二乘支持向量机的故障诊断模型

提出一种基于舍一交叉验证优化最小二乘支持向量机(LS-SVM)的旋转机械故障诊断模型。首先将故障信号EMD分解为平稳IMF分量,再选择表征故障调制特征的IMF分量并构造瞬时幅值欧式范数作为故障特征矢量输入到舍一交叉验证(leave-one-outcross-validation, LOO-CV)优化线性核LS-SVM中进行故障识别。EMD分解可自适应分离故障调制信号;瞬时幅值欧式范数矢量的不同表征各类故障的差异;舍一交叉验证优化惩罚因子可以使线性核LS-SVM克服对故障类型与模式编号映射关系先验知识的依赖,提高LS-SVM的故障预测精度和自适应诊断能力。一个深沟球轴承故障诊断实例说明该模型的有效性。     

汽轮机热耗率多模型建模方法研究

针对汽轮机热耗率难以准确计算的问题,提出了核模糊c均值与混合蛙跳算法优化最小二乘支持向量机(LS-SVM)的汽轮机热耗率多模型建模方法,用来计算不同工况下的热耗率。该方法利用核模糊c均值算法对热耗率数据聚类,采用5折交叉验证平均误差作为LS-SVM参数选择的适应度值,利用混合蛙跳算法优化参数并建立局部模型,采用开关切换得到模型输出,以此实现热耗率的多模型建模。与单一的LS-SVM模型和BP网络热耗率预测模型比较,结果表明该多模型方法有更高的预测精确和更好的泛化能力,能更准确地计算汽轮机热耗率。     

在线煤浆灰分测量的煤泥浮选优化控制系统研究

分析了煤泥浮选自动控制的特点和关键技术。设计并实现了非接触式的同位素煤浆灰分检测装置及多流分析装置，能同时检测原、精、尾煤的灰分。研制了以实时煤浆灰分为反馈量的煤泥浮选优化控制系统，用模糊控制方法调节浮选加药量和浮选槽液位，在保证精煤质量的前提下提高精煤产率。通过现场实践证明了技术路线的可行性。     

基于粒子群优化算法的金刚石砂轮磨粒边缘提取

为了实现对金刚石砂轮磨粒边缘的有效提取,将基于粒子群优化算法的Canny算子应用在金刚石砂轮磨粒的边缘检测上。用最大类间方差作为目标函数,优化Canny算子的阈值,实现边缘的有效提取。分别对实测的单颗磨粒和多颗磨粒进行了边缘提取,实验结果显示该算法可以较好地提取金刚石砂轮磨粒边缘。该方法不需要人为设定阈值,可以实现阈值的自动获取和优化。最后,利用四连通成分和八连通成分与像素总数的比值,将阈值可优化设定的Canny算子与传统的Canny算子以及最大类间方差的方法做对比,结果表明所应用的方法有效地提高了检测的准确性。     

基于最小二乘支持向量机的SOFC电特性建模

针对现有的固体氧化物燃料电池(SOFC)模型过于复杂的弊端,提出了一种基于最小二乘支持向量机(LS-SVM)的建模方法,用具有径向基函数(RBF)核函数的LS-SVM建立了SOFC电堆的非线性模型.应用仿真对建模的有效性和精度进行了检验,并与径向基函数神经网络(RBFNN)模型的辨识效果进行了比较.仿真结果证明,与RBFNN模型相比,LS-SVM模型具有较高的预测精度,这表明用LS-SVM对SOFC电堆进行建模是可行的.该LS-SVM模型的建立,对SOFC系统控制策略的研究具有一定的实用价值.     

关于降低浮选药剂消耗的研究探索

本文根据淮北选煤厂南区浮选车间在浮选生产实践中的药剂使用和操作方法探索，不断发现生产中药剂使用状况和改进方法，总结了降低浮选药剂消耗的方法措施。     

An initial alignment technique for strapdown inertial navigation systems based on large misalignment angles

针对捷联惯性导航系统( SINS)大失准角初始对准情况下非线性模型线性化导致模型不准确和影响对准精度的问题,设计了一种基于最小二乘支持向量机( LS-SVM)的大失准角对准算法.该方法采用基于加性四元数误差( AQE)的大失准角误差方程,采用简化的无迹卡尔曼滤波器(UKF)来模拟LS-SVM训练样本.捷联惯性导航系统和全球定位系统(GPS)的速度和位置误差作为LS-SVM的输入样本,简化UKF得到的失准角经小波去噪后作为输出样本.LS-SVM算法采用交叉验证法选择最佳的核函数参数.仿真结果表明,在大失准角下LS-SVM算法在对准时间和对准精度上与简化UKF和EKF相比均表现出较好的性能.     

基于AR自相关峰态值的一类轴承故障检测方法

针对轴承故障检测系统中异常样本数据不易收集以及异常样本数据分布不均导致传统分类算法出现过适应现象等现实应用问题,提出了一种基于自回归(AR)模型自相关系数峰态特征的一类故障检测方法.该方法利用正常样本生成AR模型参数,其他样本在该模型的投影形成残差序列,计算残差序列的自相关系数并取其峰态特征作为相似性的度量.实验结果表明该方法能有效地克服以AR模型参数为特征计算复杂度高且检测性能易受样本大小影响的不足.同时,文章给出了单一故障诊断模型并提出基于粒子群优化算法的阈值设定决策方法.实验中将本方法同其他以AR模型为特征的多层感知机(MLP)及自组织映射(SOM)方法进行比较,实验结果验证了本文建议方法的正确性和有效性.     

基于机器视觉的彩色印刷品套印误差检测方法研究

为了实现套印误差在线实时检测,提出了一种基于测控条的套准误差计算方法。首先将测控条进行定位,消除角度偏差;然后,基于测控条本身的特点,计算实地块边缘相对于基准色块边缘在水平和垂直方向的平移量,将此平移量与设定的标准距离相比较,即可得到其它各色印版相对于基准印版的套准偏差。实验结果表明,该算法的检测精度和效率完全满足彩色印刷品套准误差检测的要求。     

基于FMEA和LS–SVM的卷烟卷接过程质量监测与诊断

目的 针对烟支卷接过程质量监测精度低和效率差的问题,提出一种基于最小二乘支持向量机的卷接过程质量监控潜在失效模式及影响分析方法。方法 首先采用FMEA技术对卷接过程潜在失效模式进行识别和措施优先级判定。其次,通过关联度分析法得到关键失效模式的特征信号。最后,利用LS–SVM分类模型构建过程质量监测与诊断模型。通过实际生产数据对所提方法的性能进行验证。结果 对7种不同失效模式的识别,文中所提方法的总体平均识别精度达到93.53%,在识别准确性和识别效率上显著优于BPNN和SVM诊断方法,为卷烟制造过程诊断提供了新的思路。     

Optimization of reagent dosages with the use of response surface methodology and evaluation of test results with upgrading curves in graphite flotation

The aim of this study is to reduce reagent consumptions in graphite flotation with the addition of Aero 3477 promoter to the flotation stage. A three-level Box–Behnken design is used for optimization and modeling of flotation stages. For this purpose, the flotation experiments are performed in two steps. The maximum weight recovery and the minimum ash content values of the rougher concentrate are calculated as 21.41 and 32.49% using diesel oil dosage (714.24?g/t), methyl isobutyl carbinol (MIBC) dosage (212.81?g/t), and sodium silicate (Na₂SiO₃) dosage (1451.189?g/t), respectively, in the preliminary experiment tests. Subsequently, keeping the Na₂SiO₃ dosage constant, Aero 3477 promoter is added to the flotation stage to provide reduction at reagent consumptions. As a result, the reagent consumptions of diesel oil dosage (162.10?g/t), MIBC dosage (129.58?g/t), and Aero 3477 promoter dosage (168.96?g/t) are supplied to reach the approximate weight recovery and the ash content as in primary experimental results. Considerable dosage decreases in both diesel oil and MIBC are achieved using promoter. Moreover, experimental studies are also evaluated using upgrading curves.     

基于改进YOLOv5的试剂卡印刷缺陷检测算法

目的 针对试剂卡生产企业采用人工分选印刷缺陷的试剂卡存在效率低、成本高、易漏检的问题，提出一种基于深度神经网络YOLOv5s的改进试剂卡印刷缺陷检测算法YOLOv5s-EF。方法 通过图像预处理算法获得高质量的缺陷图像数据集，在YOLOv5s的主干特征提取网络中添加高效通道注意力（Efficient Channel Attention, ECA）机制，增强特征图中重要特征的表示能力；引入焦点损失函数（Focal Loss）来缓解正负样本不均衡的影响；结合印刷区域的定位结果，二次精确定位并构建方位特征向量，提出一种特征向量相似度匹配方法。结果 实验结果表明，本文提出的试剂卡印刷缺陷检测算法在测试集上的检测平均准确度可以达到97.3%，速度为22.6帧/s。结论 相较于其他网络模型，本文提出的方法可以实现对多种印刷缺陷的识别与定位，模型具有较好的检测速度和鲁棒性，有利于提高企业生产的智能化水平。     

Analysis of reverse cationic iron ore fines flotation using RSM-D-optimal design – An approach towards sustainability

Froth flotation is an important process to recover iron values from slimes, ore fines to meet the rapidly growing demand of quality iron ore concentrate for environmental and economic advantages. In the present study, responses of ?75?µm iron ore fines to flotation were analysed at different variable parameters. The D-optimal design of response surface methodology was adopted to evaluate the effect of different reagents such as collector, frother, depressant, dispersant and pH on the recovery and grade. The RSM-D optimal predicted responses with the coefficient of determination (R²) 0.97 for recovery and 0.98 for grade of iron concentrate collected from tailing launder. The optimal conditions were collector?=?0.21?kg/ton, frother?=?0.10?kg/ton, depressant?=?0.61?kg/ton, disperssant?=?0.20?kg/ton, and pH?=?11 with 64.13% Fe and 84.37% iron recovery respectively with a model desirability of 84.3%. Flotation performance was optimized by carrying out statistical design of experiments to identify main and international effects on maximum recovery and grade. This study showed that RSM-D-optimal could efficiently be applied for modeling of iron ore reverse flotation and it is very economical to obtain maximum amount of information in less interval of time with few numbers of experiments. Successful upgradation of low-grade iron ore fines through reverse flotation route with proposed reagent regimes is envisaged.     

Estimating Military Aircraft Cost Using Least Squares Support Vector Machines

A multi-layer adaptive optimizing parameters algorithm is developed for improving least squares support vector machines (LS-SVM), and a military aircraft life-cycle-cost (LCC) intelligent estimation model is proposed based on the improved LS-SVM. The intelligent cost estimation process is divided into three steps in the model. In the first step, a cost-drive-factor needs to be selected, which is significant for cost estimation. In the second step, military aircraft training samples within costs and cost-drive-factor set are obtained by the LS-SVM. Then the model can be used for new type aircraft cost estimation. Chinese military aircraft costs are estimated in the paper. The results show that the estimuted costs by the new model are closer to the true costs than that of the traditionally used methods.     

Fault detection and diagnosis method for cooling dehumidifier based on LS-SVM NARX model

Developing fault detection and diagnosis (FDD) for the cooling dehumidifier is very important for improving the equipment reliability and saving energy consumption. Due to the precise mathematic physical model for cooling dehumidifier FDD is difficult to build, a novel Nonlinear Autoregressive with Exogenous (NARX) method for the cooling dehumidifier FDD based on Least Squares Support Vector Machine (LS-SVM) is proposed. Firstly, the dehumidifier system is divided into two level models. Secondly, the parameters of the NARX model are identified by LS-SVM, and the parameters C and σ of the LS-SVM are optimized by adaptive genetic algorithm (AGA) in order to improve the model building precision. Lastly, two faults in condenser and compressor component are diagnosed by the built models. The experiment result indicates this proposed method is effective for cooling dehumidifier FDD and the model generalization ability is favorable.     

Structural damage diagnosis using incomplete static responses and LS-SVM

This paper presents novel approach to structural damage detection and estimation using incomplete static responses of a damaged structure and least squares support vector machine (LS-SVM). The presented method is based on the reduced stiffness matrix to formulate incomplete static responses as input parameters to the LS-SVM. The presented method is applied to a plane steel bridge, a four-span continuous beam and four-storey plane frame containing several damages. Also, the effect of the discrepancy in stiffness between the finite element model and the actual tested system has been investigated. The results show that the presented method is sensitive to the location and severity of the structural damage in spite of the incomplete noisy data and modelling errors.     

Degradation of 4-nitrophenol in aqueous medium by electro-Fenton method

The degradation of 4-nitrophenol by electro-Fenton (E-Fenton) method was carried out in batch recirculation mode. The effect of operating conditions such as electrical current, Fenton's reagent dosage, Fe(II) to H(2)O(2) molar ratio, and H(2)O(2) feeding time on the efficiency of E-Fenton process was investigated. It was found that E-Fenton method showed the synergetic effect on COD removal. The increase of Fenton's reagent dosage, Fe(II) to H(2)O(2) molar ratio, and the electrical current would lead to the increase of COD removal efficiency. Continuous addition of hydrogen peroxide was more effective than the addition of hydrogen peroxide in a single step and there existed an optimal H(2)O(2) feeding time for COD removal. The reaction system was modeled as a plug flow reactor (PFR) in series with a continuous stirred tank reactor (CSTR), and the pseudo-first order rate constant of COD removal was determined from the model based on the experimental data.     

Effects of type and dosages of organic depressants on pyrite floatability in microflotation system

Pyrite is considered as a disturbing sulfide mineral in the beneficiation of sulfide minerals and coal and in order to prevent floating of pyrite in the flotation, different inorganic, organic and biological depressants could be used. The most important organic pyrite depressants are starch, dextrin, guargum, carboxymethyl cellulose (CMC), ethylene diamine tetra acetic acid (EDTA), Diethylenetriamine (DETA). In this paper, flotation experiments were carried out by modified Halimond tube. At first, flotation experiments were performed to obtain optimal pyrite flotation condition (suitable collector and frother for maximum recovery). In optimal condition (collector: KAX?+?KIPX , 16?ppm; collector condition time: 4?min; frother: Poly Propylene Glycol (A65), 20?ppm; frother condition time: 30?s), pyrite recovery increased to 95.63%. Flotation kinetics experiments were performed and various kinetic models were evaluated on pyrite flotation. The results showed that flotation kinetics of pyrite is second-order. After obtaining optimal condition for flotation and kinetic studies, the effects of type and dosage of different organic depressants including starch, dextrin, guargum, carboxymethyl cellulose, EDTA and DETA at three concentrations (50, 100, 150?ppm) at neutral pH were comprehensively investigated. The results of the experiments showed that DETA can significantly reduce the recovery of pyrite at 150?ppm by 44%.     

A data mining approach to process optimization without an explicit quality function

In process optimization, the setting of the process variables is usually determined by estimating a function that relates the quality to the process variables and then optimizing this estimated function. However, it is difficult to build an accurate function from process data in industrial settings because the process variables are correlated, outliers are included in the data, and the form of the functional relation between the quality and process variables may be unknown. A solution derived from an inaccurate function is normally far from being optimal. To overcome this problem, we use a data mining approach. First, a partial least squares model is used to reduce the dimensionality of the process and quality variables. Then the process settings that yield the best output are identified by sequentially partitioning the reduced process variable space using a rule induction method. The proposed method finds an optimal setting from historical data without constructing an explicit quality function. The proposed method is illustrated with two examples obtained from steel making processes. We also show, through simulation, that the proposed method gives more stable results than estimating an explicit function even when the form of the function is known in advance.