基于FUZZY ARTMAP的加氢裂化分馏塔MIMO软测量

研究了一类多输入多输出 (MIMO)系统的软测量问题 ,将FuzzyARTMAP网络应用于加氢裂化分馏塔产品质量估计软测量 ,经实际过程数据验证指出此算法具有较强的分类及非线性多维映射能力 ,结合提出的多变量模糊PID在线校正算法 ,使所建软测量模型在线应用时具有一定的随工况变化不断校正的能力 .     

基于Fast-RVM的在线软测量预测模型

生化需氧量(biochemical oxygen demand,BOD)是评价水质好坏和污水处理效果的关键指标之一。由于污水生化处理过程复杂,在线仪表维护困难,生化需氧量无法得到快速精确地测量。针对这一问题,提出了一种基于Fast-RVM的在线软测量回归模型来实时在线预测出水指标BOD。该模型采用基于贝叶斯框架的相关向量机来在线预测输出指标,并且引入快速边际似然算法来加快模型的更新速度。通过污水数据的仿真实验,结果表明该在线模型的预测精度高于离线模型,泛化能力强,模型在线更新的快速性尤为突出,能较好地实现污水处理中出水水质的实时在线预测。     

基于不平衡学习的集成极限学习机污水处理故障诊断

污水处理过程的故障诊断数据具有高度不平衡性,影响了故障诊断效果,尤其是降低故障类别的识别正确率,导致出水水质不达标、运行费用增高和环境二次污染等问题出现。据此提出一种基于加权极限学习机集成算法的污水处理故障诊断建模方法。该方法将不平衡分类评价指标G-mean引入以加权极限学习机为基分类器的AdaBoost集成分类模型,定义新的基分类器初始权值矩阵更新规则和集成权重计算公式,用于基分类器的迭代学习。由仿真实验结果可知,基于加权极限学习机集成算法的污水处理故障诊断模型,可有效提高分类性能G-mean值和整体分类精度,特别提高了故障类的识别正确率,验证了基于加权极限学习机的集成算法在不平衡性污水处理故障诊断问题上的有效性。     

Model-based control of titer and glycosylation in fed-batch mAb production: Modeling and control system development

Therapeutic monoclonal antibodies (mAbs) are typically manufactured using mammalian cell cultures in fed-batch bioreactors, with increasing emphasis on meeting productivity and product quality attribute targets that depend strongly on such process variables as nutrient feed rates and bioreactor operating conditions. In this article, we identify, categorize, and address the challenges of achieving both productivity and product quality goals simultaneously, by developing a multivariable, model-based control system that can satisfy multiple production objectives in a fed-batch cell culture process. Here, we discuss model development and present theoretical concepts of observability and controllability that are essential to understanding and handling effectively these intrinsic challenges. Subsequently, we evaluate via simulation the performance of the outer-loop model predictive control and demonstrate the overall capability to satisfy complex production objectives in a laboratory scale bioreactor, as a first step toward the ultimate goal of creating an advanced control system for fed-batch mAb manufacturing processes.     

基于流形正则化半监督学习的污水处理操作工况识别方法

污水处理过程容易受外界冲激扰动影响,引发污泥上浮、老化、中毒、膨胀等故障工况,导致出水水质质量差,能源消耗高等问题,如何快速准确识别污水操作工况故障至关重要。针对污水工况识别过程中现有监督学习方法未利用大量未标记数据蕴含的丰富操作工况信息,采用基于流形正则化极限学习机的半监督学习方法,监视生化污水处理过程操作运行工况。该方法在学习过程中,在标记和未标记数据输入空间构建图拉普拉斯算子,通过随机特征映射建立隐含层,在流形正则化框架下,求解隐含层和输出层之间的权重,保留随机神经网络的计算效率和泛化性能。仿真实验结果表明,基于半监督极限学习机的污水处理工况识别在准确率与可靠性方面相对优于基本极限学习机方法。     

基于改进Bagging算法的高斯过程集成软测量建模

为提高对工况复杂的工业过程进行软测量建模的模型精度和泛化能力,提出了一种基于改进Bagging算法的高斯过程集成软测量建模方法。该算法采用高斯过程回归算法建立集成学习模型的基学习器,并在Bagging算法对训练样本重采样生成基学习器训练子集的基础上,采用基于正则化互信息的特征排序指标进行基学习器的输入特征抽取,实现有监督的特征扰动,从而改善学习器的差异度。待测样本进行软测量估计时,根据各高斯过程基学习器输出的方差自适应地选择基学习器进行集成输出。采用工业双酚A生产装置反应器的现场数据建模仿真,结果表明该方法是有效的。     

支持向量机在粉煤灰分类中的应用

针对以往粉煤灰分类的片面性,选择粉煤灰有关的化学成分和物理性质为特征向量,提出了基于支持向量机的粉煤灰科学分类。结果表明,此方法不仅结构简单,而且技术性能尤其是泛化能力与神经网络相比有明显提高,较全面地反映了粉煤灰的活性,从而为粉煤灰的综合利用及优化配置奠定了基础。     

Learning dynamic fault models based on a fuzzy set covering method

The synthesis of a real-time diagnostic expert system to monitor plant performance and identify faults in the event of process failure as well as signal potential failures is described. The crucial features which have been included are system reliability, use of dynamic trends in data and parameters to identify problem as well as the ability to handle complex knowledge. The emphasis here is on how a dynamic simulator is used to incorporate a learning algorithm based on a fuzzy set covering method for formulating a knowledge model of the operational characteristics, which enables debugging and testing of the system to be carried out continuously. The procedure is illustrated by reference to the operation of refinery crude oil distillation columns.     

Modeling and control of Wiener-type processes

We develop a simple relay feedback method to identify Wiener-type nonlinear processes. It separates the identification problem of the nonlinear static function from that of the linear dynamic subsystem to simplify the identification procedure significantly. Owing to the separation, the unmeasurable output of the linear dynamic subsystem can be obtained in a straightforward manner. Then, determining the model structure of the nonlinear static function becomes very simple and the estimates are robust to additive output noises. We can identify the whole activated region of the nonlinear static function as well as the ultimate information of the linear dynamic subsystem from only one relay feedback test. More information on the linear dynamic subsystem can be estimated by well-established linear system identification methods from additional tests. We use a nonlinear control strategy to compensate the nonlinear dynamics of the Wiener process so that the design parameters can be determined by usual tuning rules developed for linear processes and a high control performance can be achievable as in linear processes.     

基于变异CPSO算法的LSSVM蒸发过程软测量

在分析混沌粒子群优化算法(CPSO)和最小二乘支持向量机(SVM)理论基础上,以某氧化铝厂蒸发过程为对象,采用带有末位淘汰机制的混沌粒子群优化算法优化支持向量机的参数,建立了基于变异CPSO算法的LS-SVM的氧化铝蒸发过程软测量模型,并与PSO-LSSVM、LSSVM模型比较,研究表明,ICPSO-LSSVM模型预测准确,泛化性能好,且该模型预测结果中相对误差小于5%的样本达到92.5%,最大相对误差仅为8.1%,均方差MSE为0.05153,模型具有较高的精度,其现场实施结果表明基本可以实现出口浓度的实时在线预估。     

D-vine copulas混合模型及其在故障检测中的应用

过程监控技术是保证现代流程工业安全平稳运行及产品质量的有效手段。传统的过程监控方法大多采用维度约简方法提取数据特征,且要求过程数据必须服从高斯分布、线性等限制条件,对复杂工况条件下发生的故障难以取得较好的检测效果。因此,提出了混合D-vine copulas故障诊断模型,在不降维的情况下直接刻画数据中存在的复杂相关关系,构建过程变量的统计模型实现对存在非线性与非高斯性过程的精确描述。通过EM算法和伪极大似然估计优化混合模型参数,然后结合高密度区域（HDR）与密度分位数法等理论,构建广义贝叶斯概率（GBIP）指标实现对过程的实时监测。数值例子及在TE过程上的仿真结果说明了该混合模型的有效性及在故障检测中的良好性能。     

Determination of steel quality based on discriminating textural feature selection

Image data can be acquired from a product surface in real time by image sensor systems in chemical plants. For quality determination based on these image datasets, effective texture classification methodology is essential to handle highly dimensional images and to extract quality-related information from these product surface images.Wavelet texture analysis is useful for reducing the dimension and extracting textural information from images. Although wavelet texture analysis extracts only textural characteristics from images, the extracted features still contain unnecessary information for classification. The texture analysis method can be improved by retaining only class-dependent features and removing common features. In previous works, best basis and local discriminant basis are the most popular techniques for selecting an important basis from the wavelet packet basis. However, feature selection based on wavelet texture analysis has been studied for texture classification. Because previous methods are designed for wavelet coefficients with features for analysis, their performance is poor with wavelet texture analysis.We propose a novel texture classification methodology for quality determination based on feature selection using wavelet texture analysis. The proposed methodology applies the sequential forward floating selection (SFFS) algorithm as a feature selection strategy to select discriminating wavelet signatures using wavelet texture analysis. The proposed methodology is validated through quality determination for industrial steel surfaces. The results show that the proposed method has fewer classification errors with fewer number of features than previous methods.     

具有输入数据注意力机制的卷积神经网络用于氟化工产品质量预测

在氟化工等复杂的化工过程中,具有不同时间尺度的时变特性同时存在并作用于系统运行。这类复杂的强时变特性严重制约着各种先进控制策略的广泛应用。为了克服关键质量变量测量滞后所带来的不利因素,进一步提高氟化工过程先进控制系统的控制精度,本文提出了一种具有输入数据注意力机制的卷积神经网络（ACNN）并用于产品质量预测。通过引入注意力机制自适应地提取不同时间跨度输入数据的时间特性,来克服常规卷积神经网络因输入数据窗口固定而无法充分利用各类时变尺度特性的弊端,从而更为精准地提取氟化工过程复杂的强时变特性,更加准确地预测产品质量,辅助工业生产。应用氟化工过程真实数据和TE（Tennessee Eastman）模拟数据验证了方法的有效性和泛化性,结果表明对于强时变或同时具有长时间跨度的漂移波动而言,ACNN的质量预测模型具有更高的可靠性。     

广西鹿寨400 kt/a硫酸装置生产考核评述

介绍了广西鹿寨 40 0kt/a硫酸装置的工艺流程和概况 ,全面阐述了 72h生产考核期间的产品质量和产量、技术经济指标、原材料消耗、主要工艺指标及三废排放等方面的测定结果 ,以及关键设备的性能和运行情况。通过考核认为该装置运行平稳、运行率达 10 0 % ,产品的产量和质量、原材料消耗、各项工艺指标均达到了合同值和设计要求 ,环保指标也达到了国家标准规定的限值     

Adaptive type-2 fuzzy logic control of a bioreactor

Two adaptive type-2 fuzzy logic controllers with minimum number of rules are developed and compared by simulation for control of a bioreactor in which aerobic alcoholic fermentation for the growth of Saccharomyces cerevisiae takes place. The bioreactor model is characterized by nonlinearity and parameter uncertainty. The first adaptive fuzzy controller is a type-2 fuzzy-neuro-predictive controller (T2FNPC) that combines the capability of type-2 fuzzy logic to handle uncertainties, with the ability of predictive control to predict future plant performance making use of a neural network model of the nonlinear system. The second adaptive fuzzy controller is instead a self-tuning type-2 PI controller, where the output scaling factor is adjusted online by fuzzy rules according to the current trend of the controlled process. The performance of a type-2 fuzzy logic controller with 49 rules is used as reference.     

基于模糊核聚类的乙烯裂解深度DE-LSSVM多模型建模

乙烯裂解深度的建模与控制对于裂解炉的实时优化具有重要意义。针对石脑油原料组分复杂、油品特性波动大等状况,采用模糊核聚类对石脑油数据库进行最优划分,建立最小二乘支持向量机的多模型,对于最小二乘支持向量机中模型的参数选取,利用差分进化算法进行参数寻优,提高了模型的精度和泛化能力。通过对现场数据的建模实验,结果表明：基于模糊核聚类的乙烯裂解深度最小二乘支持向量机多模型跟踪性能良好,预测精度较高。     

Adaptive local kernel-based learning for soft sensor modeling of nonlinear processes

Soft sensor techniques have been widely used to estimate product quality or other key indices which cannot be measured online by hardware sensors. Unfortunately, their estimation performance would deteriorate under certain circumstances, e.g., the change of the process characteristics, especially for global learning approaches. Meanwhile, local learning methods always only utilize input information to select relevant instances, which may lead to a waste of output information and inaccurate sample selection. To overcome these disadvantages, a new local modeling algorithm, adaptive local kernel-based learning scheme (ALKL) is proposed. First, a new similarity measurement using both input and output information is proposed and utilized in a supervised locality preserving projection technique to select relevant samples. Second, an adaptive weighted least squares support vector regression (AW-LSSVR) is employed to establish a local model and predict output indices for each query data. In AW-LSSVR, instead of using traditional cross-validation methods, the trade-off parameters are adjusted iteratively and the local model is updated recursively, which reduces the computational complexity a lot. The proposed ALKL is applied to an online crude oil endpoint prediction in an industrial fluidized catalytic cracking unit (FCCU) process. The experimental results demonstrate the high precision of our ALKL approach.     

Study and Application of Fault Prediction Methods with Improved Reservoir Neural Networks

Time-series prediction is one of themajor methodologies used for fault prediction. Themethods based on recurrent neural networks have been widely used in time-series prediction for their remarkable non-liner mapping ability. As a new recurrent neural network, reservoir neural network can effectively process the time-series prediction. However, the ill-posedness problemof reservoir neural networks has seriously restricted the generalization performance. In this paper, a fault prediction algorithm based on time-series is proposed using improved reservoir neural networks. The basic idea is taking structure risk into consideration, that is, the cost function involves not only the experience risk factor but also the structure risk factor. Thus a regulation coefficient is introduced to calculate the outputweight of the reservoir neural network. As a result, the amplitude of outputweight is effectively controlled and the ill-posedness problemis solved. Because the training speed of ordinary reservoir networks is naturally fast, the improved reservoir networks for time-series prediction are good in speed and generalization ability. Experiments on Mackey-Glass and sunspot time series prediction prove the effectiveness of the algorithm. The proposed algorithm is applied to TE process fault prediction. We first forecast some timeseries obtained from TE and then predict the fault type adopting the static reservoirs with the predicted data. The final prediction correct rate reaches 81%.     

Decentralized Fault Isolation by a Combination of Transfer Entropy and Classification Methods

Early fault detection and isolation in industrial systems is vitally necessary to prevent any potential product damage. The paper proposes a new decentralized multi-unit fault isolation methodology in which all the known process faults with similar time signatures are grouped into appropriate categories. An innovative genetic algorithm-based method is introduced to explore for optimum plant zones in a large-scale plant wide search to appropriately configure each architectural unit, having less reliance on excess process variables with redundant and uncorrelated diagnostic information. The methodology employs a set of Bayes and radial basis function neural network classifiers to properly isolate the most usual known faults. A new idea based on transfer entropy algorithm has been integrated in the decentralized configuration to be triggered for isolation of novel faults which have been left unrecognized by the set of maintained classifiers. Experimental results clearly demonstrate that the proposed methods are considerably superior to the conventional centralized methods.     

A GENETIC NEURAL FUZZY SYSTEM AND ITS APPLICATION IN QUALITY PREDICTION IN THE INJECTION PROCESS

A genetic neural fuzzy system (GNFS) is presented and introduced to quality prediction in the injection process. A hybrid-learning algorithm is proposed, which is divided into two stages to train GNFS. During the first learning stage, the genetic algorithm is used to optimize the structure of GNFS and the membership function of each fuzzy term because of its capability of parallel and global search. On the basis of the first optimized training stages, the back-propagation algorithm (BP algorithm) is adopted to update the parameters of the GNFS to improve its predicting precision and reduce the computation time. The process of constructing a quality prediction model for an injection process based on GNFS is described in detail. The predicted weight of the molded part from the model based on GNFS demonstrates that the proposed GNFS has superior performance and good generalization capability in quality prediction in the injection process.