Soft sensor of chemical processes with large numbers of input parameters using auto-associative hierarchical neural network

To explore the problems of monitoring chemical processes with large numbers of input parameters, a method based on Auto-associative Hierarchical Neural Network (AHNN) is proposed. AHNN focuses on deali...     

基于GAMs和Bootstrap的生化过程建模和故障监控(英文)

Fault monitoring of bioprocess is important to ensure safety of a reactor and maintain high quality of products. It is difficult to build an accurate mechanistic model for a bioprocess, so fault monitoring based on rich historical or online database is an effective way. A group of data based on bootstrap method could be resampling stochastically, improving generalization capability of model. In this paper, online fault monitoring of generalized additive models (GAMs) combining with bootstrap is proposed for glutamate fermentation process. GAMs and bootstrap are first used to decide confidence interval based on the online and off-line normal sampled data from glutamate fermentation experiments. Then GAMs are used to online fault monitoring for time, dissolved oxygen, oxygen uptake rate, and carbon dioxide evolution rate. The method can provide accurate fault alarm online and is helpful to provide useful information for removing fault and abnormal phenomena in the fermentation.     

在同心和偏心圆环中有屈服应力的黏塑性流体流动的模型和数值模拟(英文)

Numerical solution of yield viscoplastic fluid flow is hindered by the singularity inherent to the Herschel-Bulkley model.A finite difference method over the boundary-fitted orthogonal coordinate system is util-ized to investigate numerically the fully developed steady flow of non-Newtonian yield viscoplastic fluid through concentric and eccentric annuli.The fluid rheology is described with the Herschel-Bulkley model.The numerical simulation based on a continuous viscoplastic approach to the Herschel-Bulkley model is found in poor accordance with the experimental data on volumetric flow rate of a bentonite suspension.A strict mathematical model for Herschel-Bulkley fluid flow is established and the corresponding numerical procedures are proposed.However,only the case of flow of a Herschel-Bulkley fluid in a concentric annulus is resolved based on the presumed flow structure by using the common optimization technique.Possible flow structures in an eccentric annulus are pre-sumed,and further challenges in numerical simulation of the Herschel-Bulkley fluid flow are suggested.     

A local and global statistics pattern analysis method and its application to process fault identification☆

Traditional principal component analysis (PCA) is a second-order method and lacks the ability to provide higher-order representations for data variables. Recently, a statistics pattern analysis (SPA) framework has been incor-porated into PCA model to make full use of various statistics of data variables effectively. However, these methods omit the local information, which is also important for process monitoring and fault diagnosis. In this paper, a local and global statistics pattern analysis (LGSPA) method, which integrates SPA framework and locality pre-serving projections within the PCA, is proposed to utilize various statistics and preserve both local and global in-formation in the observed data. For the purpose of fault detection, two monitoring indices are constructed based on the LGSPA model. In order to identify fault variables, an improved reconstruction based contribution (IRBC) plot based on LGSPA model is proposed to locate fault variables. The RBC of various statistics of original process variables to the monitoring indices is calculated with the proposed RBC method. Based on the calculated RBC of process variables' statistics, a new contribution of process variables is built to locate fault variables. The simula-tion results on a simple six-variable system and a continuous stirred tank reactor system demonstrate that the proposed fault diagnosis method can effectively detect fault and distinguish the fault variables from normal variables.     

Pt／C催化邻硝基氯苯加氢反应的动力学研究

The kinetics of catalytic hydrogenation of ortho-nitrochlorobenzene to 2,2‘-dichloroazoxybenzene on platinum/carbon catalyst is investigated in a slurry reactor with the temperature range of 313--343 K, and orthochloroaniline is formed as a byproduct. Models based on Rideal-Eley and Langmuir-Hinshelwood mechanism have been proposed based on the rate data and the kinetic regime. The former model can be used to fit the experimental data better. Reaction controlling steps are physical adsorption of hydrogen and adsorbed ortho-nitrochlorobenzene reacted on the surface of catalyst.     

确定多孔介质孔结构参数的渗流网络分析法

According to the simulation of nitrogen sorption process in porous media with three-dimensional network model, and the analysis for such a process with percolation theory, a new method is proposed to determine a pore structure parameter--mean coordination number of pore network, which represents the connectivity among a great number of pores. Here the “chamber-throat“ model and the Weibull distribution are used to describe the pore geometry and the pore size distribution respectively. This method is based on the scaling law of percolation theory after both effects of sorption thermodynamics and pore size on the sorption hysteresis loops are considered. The results show that it is an effective procedure to calculate the mean coordination number for micro- and meso-porous media.     

Kinetics of Ortho-nitrochlorobenzene Hydrogenation on Platinum/Carbon Catalyst

The kinetics of catalytic hydrogenation of ortho-nitrochlorobenzene to 2,2'-dichloroazoxybenzene on platinum/carbon catalyst is investigated in a slurry reactor with the temperature range of 313-343 K, and ortho-chloroaniline is formed as a byproduct. Models based on Rideal-Eley and Langmuir-Hinshelwood mechanism have been proposed based on the rate data and the kinetic regime. The former model can be used to fit the experimental data better. Reaction controlling steps are physical adsorption of hydrogen and adsorbed ortho-nitrochlorobenzene reacted on the surface of catalyst.     

基于邻域保留空间中模糊C聚类的多模态过程监控(英文)

For complex industrial processes with multiple operational conditions, it is important to develop effective monitoring algorithms to ensure the safety of production processes. This paper proposes a novel monitoring strategy based on fuzzy C-means. The high dimensional historical data are transferred to a low dimensional subspace spanned by locality preserving projection. Then the scores in the novel subspace are classified into several overlapped clusters, each representing an operational mode. The distance statistics of each cluster are integrated though the membership values into a novel BID (Bayesian inference distance) monitoring index. The efficiency and effectiveness of the proposed method are validated though the Tennessee Eastman benchmark process.     

含反应节点过程数据校正的扩展独立物流法

Among the techniques developed for bilinear data reconciliation problems, the method based on independent flows is well known in terms of both accuracy and efficiency. However, the independent flow method is not effective when reactor units are present in the process. In this paper, an extended independent flow method is proposed for the data reconciliation of the process with chemical reaction. By the new method, the independent flows finding algorithm is adjusted to avoid the difficulties caused by the reactors in the process, and the reaction constraints are introduced into the objective function of data reconciliation. As a result, the new method can be applied to the process with chemical reaction while retaining high solution accuracy. To test the performance, the new method and the most typical Crowe‘s projection method are used in the data reconciliation of a typical industrial process. The results show that the new method can effectively accomplish the data reconciliation of the muhicomponent process with chemical reaction and gives more accurate estimates than the Crowe‘s method.     

规整填料内气液两相流动与传质的计算传质学模型

Detailed investigation of flow behavior in structured packing distillation columns is of great importance in accurate prediction of process efficiency and development of more efficient and optimal equipment internals. In this study, a three-dimensional two-phase flow model based on VOF method for simulating the hydrodynamics and mass-transfer behavior in a typical representative unit of the structured packing is developed. In the proposed model, the model is used for the closure of turbulent mass transfer equation. By solving the proposed model, the velocity distribution, phase fraction profile and concentration field are obtained. Using these data, the total liquid holdup, the wetted area and the separation efficiency [height equivalent to a theoretical plate (HETP)] are estimated. For testing the model validation, the simulated HETPs are compared with our previous experimental data obtained in a 150 mm-diameter column containing Mellapak 350Y operating at the pressures of 0.6-1.8 MPa. The compari-son shows that they are in satisfactory agreement, with an average absolute deviation (AAD) of 25.4%.     

基于数据属性划分的递阶ELM研究及化工应用

针对极限学习机（ELM）不能有效处理化工过程中强耦合、带噪声的高维数据建模问题,提出了一种基于数据属性划分的递阶ELM神经网络DHELM。该神经网络采用数据属性划分（DAD）方法对高维输入进行聚类、建立自联想子网,并将自联想子网所提取的特征分量作为极限学习机的输入进行建模。同时,利用UCI标准数据集进行了测试,通过工业应用实例进行了验证,并进行了模型对比。结果表明,DHELM网络在处理复杂高维数据时具有收敛速度快、建模精度高、网络稳定性强的特点,为神经网络发展及其化工应用提供了新思路。     

基于人工神经网络的粉煤灰科学分类

鉴于目前国内在粉煤灰分类方面存在的片面性,我们结合人工神经网络(ANN)理论,提出了基于人工神经网络的粉煤灰科学分类方法。该方法充分考虑了粉煤灰的火山灰活性,并考虑了细度、玻璃体、烧失量、K2O、SO3、CaO多种因素对分类的影响。实例表明,按本文提出的方法建立的网络模型比较合理,精度比较高,克服了以往分类方法的片面性,能比较全面地反映粉煤灰的品质性能,为粉煤灰的多元化利用提供依据。     

A novel fault classification method using reconstructed distance-based discriminant locality preserving projection for industrial processes

Fault detection and classification is a crucial issue in modern industrial processes for ensuring steady operation and high product quality. The process data collected and stored fully reflect the equipment running state and the production process. Moreover, the extracted nonlinear features can directly affect the effectiveness of the data-driven fault classification model. In this paper, a novel fault classification method based on nonlinear feature extraction using reconstructed distance-based discriminant locality preserving projection (RD-DLPP) is proposed. First, a hypersphere model for each class of data is developed according to the spatial structures and classes information in high-dimensional space. The hyperspheres are used as indicators to evaluate the discriminatory difficulty of samples. Second, the constraints of the correlations between the k-nearest neighbour points of the sample and the hypersphere are introduced, which can efficiently reconstruct new measure metrics between the sample and its k-nearest neighbour points. Finally, an improved fault classification model based on RD-DLPP is established for the construction of the highly discriminant subspace. The Bayesian decision is then used to classify the samples. The feasibility and efficiency of the proposed method are verified by the Tennessee Eastman process as a case study.     

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.     

Haemoglobin response modelling under erythropoietin treatment: Physiological model-informed machine learning method

Patients with renal anaemia are usually treated with recombinant human erythropoietin (EPO) because of insufficient renal EPO secretion. The establishment of a good haemoglobin (Hgb) response model is a necessary condition for dose optimization design. The purpose of this paper is to apply physics-informed neural networks (PINN) to build the Hgb response model under EPO treatment. Neural network training is guided by a physiological model to avoid overfitting problems. During the training process, the parameters of the physiological model can be estimated simultaneously. To handle differential equations with impulse inputs and time delays, we propose approximate model equations for the pharmacokinetic (PK) model and the pharmacodynamic (PD) model, respectively. The modified PK/PD model was incorporated into PINN for training. Tests on simulated data and clinical data show that the proposed method has better performance than data-driven modelling methods and the traditional physiological modelling based on the least squares method.     

Walsh Fourier Transform of Locally Stationary Time Series

A new time-frequency model and a method to classify time series data are proposed in this article. By viewing the observed signals as realizations of locally dyadic stationary (LDS) processes, a LDS model can be used to provide a time-frequency decomposition of the signals, under which the evolutionary Walsh spectrum and related statistics can be defined and estimated. The classification procedure is as follows. First choose a training data set that comprises two groups of time series with a known group. Then compute the time frequency feature (the energy) using the training data set, and use a best tree method to maximize the discrepancy of this feature between the two groups. Finally, choose the testing data set with the unknown group as validation data, and use a discriminant statistic to classify the validation data to one of the groups. The classification method is illustrated via an electroencephalographic dataset and the Ericsson B transaction time dataset. The proposed classification method performs better for integer-valued time series in terms of classification error rates in both simulations and real-life applications.     

BPNN鲁棒评价函数

在分析过拟合现象产生原因的基础上,总结前人研究成果,提出了构造鲁棒评价函数的基本思想和充分条件.据此建立了两个鲁棒评价函数,实例计算表明,它比最小二乘评价函数能更有效地减少过拟合现象.     

An Artificial Neural Network approximation based decomposition approach for parameter estimation of system of ordinary differential equations

In this work a new approach for parameter estimation which is based upon decomposing the problem into two subproblems is proposed, the first subproblem generates an Artificial Neural Network (ANN) model from the given data and then the second subproblem uses the ANN model to obtain an estimate of the parameters. The analytical derivates from the ANN model obtained from the first subproblem are used for obtaining the differential terms in the formulation of the second subproblem. This greatly simplifies the parameter estimation problem. The key advantage of the proposed approach is that solution of a large optimization problem requiring high computational resources is avoided and instead two smaller problems are solved. This approach is particularly useful for large and noisy data sets and nonlinear models where ANN models are known to perform quite well and therefore plays an important role in the solution of the overall parameter estimation problem.     

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。