基于混合Copula模型的铝电解槽多参数相关性分析

针对铝电解大型预焙槽操作参数较多且彼此耦合性强,难以进行准确的概率分布描述和相关性分析问题,提出一种基于混合Copula模型的铝电解槽多参数相关性分析方法。在铝电解槽参数分布类型未知的情况下,首先利用非参数核密度函数估计建立变量的边缘密度函数;再构建基于混合Copula模型的多变量联合分布函数,并通过权重参数调节不同类型Copula函数的贡献比重;最后利用极大似然法对模型参数进行估计。对取自某厂170 kA铝电解槽的1824组真实样本数据进行实验,结果得到的3种距离指标分别是0.3169、0.6239和0.9276,均优于其他单一Copula函数,表明本方法是对超低电压下具有非稳态非均一特征的多参数进行相关性分析的一种有效途径。     

基于混合Copula模型的铝电解槽多参数相关性分析

针对铝电解大型预焙槽操作参数较多且彼此耦合性强,难以进行准确的概率分布描述和相关性分析问题,提出一种基于混合Copula模型的铝电解槽多参数相关性分析方法。在铝电解槽参数分布类型未知的情况下,首先利用非参数核密度函数估计建立变量的边缘密度函数;再构建基于混合Copula模型的多变量联合分布函数,并通过权重参数调节不同类型Copula函数的贡献比重;最后利用极大似然法对模型参数进行估计。对取自某厂170 kA铝电解槽的1824组真实样本数据进行实验,结果得到的3种距离指标分别是0.3169、0.6239和0.9276,均优于其他单一Copula函数,表明本方法是对超低电压下具有非稳态非均一特征的多参数进行相关性分析的一种有效途径。     

基于阳极电流波动的铝电解槽槽况诊断系统

引言 铝电解槽在正常生产过程中处于高温状态,且其内的高温熔体具有很强的腐蚀性,一般的材料在其中会很快被腐蚀.因此铝电解槽的工作状态很难直接探测并实现在线显示,但铝电解槽工作状态参数的获得对槽况在线识别和操作优化是至关重要的.     

不同进电方式对铝电解槽阳极电流非接触式测量的影响

针对目前铝电解槽阳极电流在线测试方法的不足,基于铝电解槽电磁场的理论,借鉴国外学者通过测试导杆附近的磁场来得到阳极电流的思想,提出了一种阳极电流非接触式测试方法,并着重针对铝电解槽不同的进电方式分析了阳极电流的测量误差,并对误差进行了修正,对测试方案进行了优化,旨在实现阳极电流的实时监测和在线控制。研究结果表明,300 kA大面五点进电槽的阳极电流测量误差可以控制在±0.25%之内,而400 kA大面六点进电槽、200 kA大面四点进电槽的阳极电流测量误差分别达到了±0.8%和±3%;可见不同进电方式对阳极电流的测量误差有较为明显的影响,但最大误差均不超过3%。因此,对于不同进电方式及结构的铝电解槽阳极电流的在线监测,该非接触式测试方法均具有较好的适用性。     

基于主元分析-概率神经网络的制冷系统故障诊断

制冷系统由于内部物质形态的多样性以及系统参数间的高度耦合而较为复杂,也增加了出现故障后的检测及诊断难度。针对制冷系统常见的7种故障,包括局部故障与系统故障,运用主元分析法提取故障样本主要特征,对样本进行降维处理后,基于概率神经网络进行故障诊断。主元分析法可将原始的62个参数分解为相互独立的主元,根据累计贡献率选取一定量的主元,并将其样本输入概率神经网络进行故障诊断,结果表明结合主元分析后的概率神经网络在一定范围内对spread值不敏感,不仅诊断正确率有所提高,而且缩短了诊断耗时。可见,主元分析法的使用可有效优化概率神经网络的诊断性能。     

大型预焙铝电解槽与炭阳极

铝电解槽大型化对炭阳极的质量提出了新的要求。大型槽运行的实践证明，炭阳极的质量不仅直接影响着铝电解槽的运行状况及其技术经济指标，而且长期使用低质量炭阳极将会导致铝电解槽物料过耗、炭渣过多、槽子过热、槽帮过空，进而引发病槽，缩短铝电解槽的寿命。因此说，阳极质量是大型预焙槽的正常运行的关键因素之一。本文就大型预焙槽的特点、阳极过耗与质量、炭渣的生成与影响、粘结剂的特性等几个方面，从铝电解及铝用炭素生产工艺的角度论述了我国炭阳极的质量现状、主要问题及改进途径。     

半导体温差发电技术在铝电解槽中的应用研究

测量铝电解槽的散热量、槽壁热流密度及温度,对半导体温差发电技术用于铝电解槽的可行性进行分析。结合铝电解槽和半导体温差发电技术的特点,设计符合铝电解槽特点的半导体温差发电装置,测量发电装置的功率和效率。测试结果显示半导体温差发电技术可用于铝电解槽余热发电。     

炭素专利

铝电解用阳极炭块底部或侧部成槽工艺;一种处理铝电解槽废槽衬的方法;活性炭制造用炭化装置;一种高强防水后成型生物质炭的生产方法;一种炭中空纤维及其合成方法和用途;     

隔膜电槽的分布电压测量方法及探讨

通过对国内隔膜法金属阳极电解槽的调查,实感对现有隔膜电槽分布电压进行测量之必要。为加快金属阳极电解槽的推广、了解、掌握槽电压的分布规律,特向读者推荐关于隔膜电解槽分布电压的测量方法,供选择、设计金属阳极电解槽槽型,提供有关参考数据。     

介绍二种铝电解槽用异型炭块

吉林炭素厂长春分厂(以下简称长春分厂),1990年为朝鲜民主主义共和国南浦、北仑铝厂试制一种铝电解槽异型炭块,这种炭块与我国日前使用铝槽炭块截然不同,需六面加工,其中五个面带有沟槽(锯齿槽、半圆槽、燕尾槽、矩形槽)。     

基于模糊神经网络的离心式压缩机故障诊断

利用模糊神经网络建立离心式压缩机故障诊断模型,使用梯度下降法进行学习训练,引用万能逼近定理设计离心式压缩机非线性系统函数逼近误差公式。仿真结果表明:出现故障后,逼近误差函数能够精确诊断离心式压缩机故障。     

Diagnosis of working conditions of an aluminum reduction cell based on wavelet packets and fuzzy neural network

In this paper, fuzzy neural network is combined with wavelet packet analysis for diagnosis of working conditions of aluminum reduction cells. The sample data is pre-processed using best wavelet packet basis for the forecast and then an adaptive-network-based fuzzy inference system (ANFIS) is established for diagnosis of working conditions. The wavelet packet analysis was used to extract the characteristic of signal according to the frequency spectrum characteristics of voltage vibration signal of aluminum reduction cells. The signals were decomposed into eight frequency bands and the information pre-conditioned was used as an energy characteristic vector. The structure of ANFIS is given and the membership function is developed according to the actual situation. All simulated working conditions are emulated on 350 KA pre-baked aluminum reduction cells. The feasibility of this novel method is proved by the simulation results.     

径向基函数神经网络在砂体厚度预测中的应用

目前,人工智能神经网络在地震储层参数的预测方面具有广泛的应用,最常用的为BP神经网络,但是效果并不是十分理想。径向基函数神经网络(RBFN)是一种前馈神经网络,其在函数逼近、模式识别方面都优于BP网络,已经在岩性识别、孔渗预测方面取得了较好的应用效果。本文首次将此方法运用于预测砂体厚度,利用地震属性信息和神经网络的学习,基于实际数据计算,最后计算出相应的砂体厚度值,并与实测值进行误差分析。实例分析表明,利用径向基函数神经网络进行砂体厚度预测具有一定的可行性和实用价值。     

ARTIFICIAL NEURAL NETWORKS IN LIQUID-LIQUID TWO-PHASE FLOW

This research investigates the use of neural networks for identification of oil-water two-phase flow pattern in a horizontal conduit. The investigation seeks to find out if neural networks are useful tools for this application. In order to find a suitable network, all of the various types of networks available are studied and appropriate networks are selected using MATLAB as a computational tool. Four networks are selected for this investigation: feed-forward back propagation (FFBP), radial based function (RBF), probabilistic neural network (PNN), and learning vector quantization (LVQ) Networks. These various networks are created and tested to evaluate the relationships between key variables and performance. Once the networks are optimized, various architectures are compared based on mean square error (MSE), construction time, complexity, and the ability to identify transition regions. PNN is found to be the best network for this particular application, followed by FFBP, RBF, and LVQ. It is observed that making a network larger or more powerful does not necessarily improve the performance. Moreover, doing so causes a network to lose its ability to generalize by eventually “over-fitting” the training data.     

Enhanced statistical analysis of nonlinear processes using KPCA, KICA and SVM

In this paper, some drawbacks of original kernel independent component analysis (KICA) and support vector machine (SVM) algorithms are analyzed for the purpose of multivariate statistical process monitoring (MSPM). When the measured variables follow non-Gaussian distribution, KICA provides more meaningful knowledge by extracting higher-order statistics compared with PCA and kernel principal component analysis (KPCA). However, in real industrial processes, process variables are complex and are not absolutely Gaussian or non-Gaussian distributed. Any single technique is not sufficient to extract the hidden information. Hence, both KICA (non-Gaussion part) and KPCA (Gaussion part) are used for fault detection in this paper, which combine the advantages of KPCA and KICA to develop a nonlinear dynamic approach to detect fault online compared to other nonlinear approaches. Because SVM is available for classifying faults, it is used to diagnose fault in this paper.For above mentioned kernel methods, the calculation of eigenvectors and support vectors will be time consuming when the sample number becomes large. Hence, some dissimilar data are analyzed in the input and feature space.The proposed approach is applied to the fault detection and diagnosis in the Tennessee Eastman process. Application of the proposed approach indicates that proposed method effectively captures the nonlinear dynamics in the process variables.     

基于Sigmoid函数参数调整的双隐层BP神经网络的板形预测

提出一种改进的BP神经网络处理板形缺陷数据的方法,建立双隐层BP神经网络模型,并对Sigmoid激活函数的形状进行调节。将其应用到冷轧的板形缺陷识别中,与利用Levenberg-Marquardt规则训练的BP神经网络预测结果作对比,表明该方法不仅有效地减少双隐层BP网络的学习时间,同时改善了网络的泛化能力,有利于板形缺陷在线识别。     

Nonlinear dimension reduction based neural modeling for distributed parameter processes

Many chemical processes are nonlinear distributed parameter systems with unknown uncertainties. For this class of infinite-dimensional systems, the low-order model identification from process data is very important in practice. The dimension reduction with a principal component analysis (PCA) is only a linear approximation for nonlinear problem. In this study, a nonlinear dimension reduction based low-order neural model identification approach is proposed for nonlinear distributed parameter processes. First, a nonlinear principal component analysis (NL-PCA) network is designed for the nonlinear dimension reduction, which can transform the high-dimensional spatio-temporal data into a low-dimensional time domain. Then, a neural system can be easily identified to model this low-dimensional temporal data. Finally, the spatio-temporal dynamics can be reproduced using the nonlinear time/space reconstruction. The simulations on a typical nonlinear transport-reaction process show that the proposed approach can achieve a better performance than the linear PCA based modeling approach.     

水泥煅烧工段机理分析与神经网络预测方法

采用神经网络的方法建立水泥预分解窑煅烧工段的预测模型。选择合理的状态与控制变量,通过采集实际运行数据来训练神经网络。构建的基于BPNN神经网络的煅烧预测模型能够较好地拟合采样数据,具有较好的泛化能力。     

一种分布式模糊神经网络在熔融指数软测量中的应用

采用Mamdani模型作为模糊分类器 ,利用神经网络建立非线性模型 ,构造一种分布式神经网络。采用多组样本数据建模 ,根据各输入模糊子集和隶属度函数 ,将输入样本空间模糊分割成多个子空间 ,对每个子空间用一个神经网络模型建立映射关系。对每一组输入向量在确定归属类后 ,自动切换至对应的子网络作为输入 ,该子网络的输出值则作为分布式网络的输出。仿真结果表明 ,该方法与用单个神经元网络相比 ,明显提高了模型的精度和泛化能力。     

基于RBF的玻璃钢天线罩逆向工程曲面重构

研究目的是通过人工神经网络方法重构玻璃钢天线罩曲面。径向基函数(RBF)神经网络具有很强的非线性逼近能力,根据给定的天线罩外表面数据及采集的天线罩内表面数据,采用RBF神经网络对玻璃钢天线罩进行逆向工程曲面重构。典型工程实例的计算结果证明,该方法的拟合、重构精度高,并且训练速度快,具有很高的实用推广价值。