动态多变量过程监控研究

Chemical process variables are always driven by random noise and disturbances. The closed-loop control yields process measurements that are auto and cross correlated. The influence of auto and cross correlations on statistical process control （SPC） is investigated in detail by Monte Carlo experiments. It is revealed that in the sense of average performance, the false alarms rates （FAR） of principal component analysis （PCA）, dynamic PCA are not affected by the time-series structures of process variables. Nevertheless, non-independent identical distribution will cause the actual FAR to deviate from its theoretic value apparently and result in unexpected consecutive false alarms for normal operating process. Dynamic PCA and ARMA-PCA are demonstrated to be inefficient to remove the influences of auto and cross correlations. Subspace identification-based PCA （SI-PCA） is proposed to improve the monitoring of dynamic processes. Through state space modeling, SI-PCA can remove the auto and cross correlations efficiently and avoid consecutive false alarms. Synthetic Monte Carlo experiments and the application in Tennessee Eastman challenge process illustrate the advantages of the proposed approach.     

基于部分DPCA的动态系统故障隔离

Principal component analysis （PCA） is a useful tool in process fault detection, but offers little support on fault isolation. In this article, structured residual with strong isolation property is introduced. Although it is easy to get the residual by transformation matrix in static process, unfortunately, it becomes hard in dynamic process under control loop. Therefore, partial dynamic PCA（PDPCA） is proposed to obtain structured residual and enhance the isolation ability of dynamic process monitoring, and a compound statistic is introduced to resolve the problem resulting from independent variables in every variable subset. Simulations on continuous stirred tank reactor （CSTR） show the effectiveness of the proposed method.     

输入训练神经网络的维数约简算法及其在化工过程建模中的应用

Many applications of principal component analysis （PCA） can be found in dimensionality reduction. But linear PCA method is not well suitable for nonlinear chemical processes. A new PCA method based on improved input training neural network （IT-NN） is proposed for the nonlinear system modelling in this paper. Momentum factor and adaptive learning rate are introduced into learning algorithm to improve the training speed of IT-NN. Contrasting to the auto-associative neural network （ANN）, IT-NN has less hidden layers and higher training speed. The effectiveness is illustrated through a comparison of IT-NN with linear PCA and ANN with experiments. Moreover, the IT-NN is combined with RBF neural network （RBF-NN） to model the yields of ethylene and propylene in the naphtha pyrolysis system. From the illustrative example and practical application, IT-NN combined with RBF-NN is an effective method of nonlinear chemical process modelling.     

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.     

Synthesis of Multi-component Mass-exchange Networks

This paper presents a superstructure-based formulation for the synthesis of mass-exchange networks （MENs） considering multiple components. The superstructure is simplified by directly using the mass separation agents （MSA） from their sources, and therefore the automatic synthesis of the multi-component system involved in the MENs can be achieved without choosing a ＇key-component＇ either for the whole process or the mass exchangers A mathematical model is proposed to carry out the optimization process. The concentrations, flow rates, matches and unit operation displayed in the obtained network constitute the exact representation of the mass exchange process in terms of all species in the system. An example is used to illustrate and demonstrate the application of the proposed method.     

基于主元分析和核密度估计的多变量统计过程监控及在工厂聚丙烯催化剂反应器的应用

Abstract Data-driven tools, such as principal component analysis （PCA） and independent component analysis （ICA） have been applied to different benchmarks as process monitoring methods. The difference between the two methods is that the components of PCA are still dependent while ICA has no orthogonality constraint and its latentvariables are independent. Process monitoring with PCA often supposes that process data or principal components is Gaussian distribution. However, this kind of constraint cannot be satisfied by several practical processes. To ex-tend the use of PCA, a nonparametric method is added to PCA to overcome the difficulty, and kernel density estimation （KDE） is rather a good choice. Though ICA is based on non-Gaussian distribution intormation, .KDE can help in the close monitoring of the data. Methods, such as PCA, ICA, PCA.with .KDE（KPCA）, and ICA with KDE,（KICA）, are demonstrated and. compared by applying them to a practical industnal Spheripol craft polypropylene catalyzer reactor instead of a laboratory emulator.     

PCA weight and Johnson transformation based alarm threshold optimization in chemical processes

To alleviate the heavy load of massive alarm on operators, alarm threshold in chemical processes was optimized with principal component analysis(PCA) weight and Johnson transformation in this paper. First, few variables that have high PCA weight factors are chosen as key variables. Given a total alarm frequency to these variables initially, the allowed alarm number for each variable is determined according to their sampling time and weight factors. Their alarm threshold and then control limit percentage are determined successively. The control limit percentage of non-key variables is determined with 3σ method alternatively. Second, raw data are transformed into normal distribution data with Johnson function for all variables before updating their alarm thresholds via inverse transformation of obtained control limit percentage. Alarm thresholds are optimized by iterating this process until the calculated alarm frequency reaches standard level(normally one alarm per minute). Finally,variables and their alarm thresholds are visualized in parallel coordinate to depict their variation trends concisely and clearly. Case studies on a simulated industrial atmospheric-vacuum crude distillation demonstrate that the proposed alarm threshold optimization strategy can effectively reduce false alarm rate in chemical processes.     

基于非线性主元分析和符号有向图的故障诊断方法

Nonlinear principal component analysis（NLPCA）fault detection method achieves good detection results especially in a nonlinear process.Signed directed graph（SDG）model is based on deep-going information,which excels in fault interpretation.In this work,an NLPCA-SDG fault diagnosis method was proposed.SDG model was used to interpret the residual contributions produced by NLPCA.This method could overcome the shortcomings of traditional principal component analysis（PCA）method in fault detection of a nonlinear process and the shortcomings of traditional SDG method in single variable statistics in discriminating node conditions and threshold values.The application to a distillation unit of a petrochemical plant illustrated its validity in nonlinear process fault diagnosis.     

Fault diagnosis for distillation process based on CNN–DAE

Distillation is the most widely used operation for liquid mixture separation in the chemical industry. It is of great importance to detect and diagnose faults in distillation process. Due to the strong feedback and coupling of processes in a distillation column, it is difficult to use deep auto-encoders(DAEs) alone to achieve good results in detecting and diagnosing faults, in terms of accuracy and efficiency. This paper proposes a hybrid fault-diagnosis model based on convolutional neural networks(CNNs) and DAEs, by integrating the powerful capability of CNN in feature extraction and of DAE in classification. A case study was carried out with the distillation process of depropanization. It is shown that the proposed hybrid model is of good performance compared to other models, in terms of the accuracy of fault detection in such a process. Also, with the increase of structural layers of the CNN–DAE model, the diagnostic accuracy will be improved, with an optimal accuracy of 92.2%.     

过程系统中波动传播辨识及源定位的集成方法研究(英文)

An integrated method for identifying the propagation of multi-loop process oscillations and their source location is proposed in this paper. Oscillatory process loop variables are automatically selected based on the component-related ratio index and a mixing matrix, both of which are obtained in data preprocessing by spectral independent component analysis. The complex causality among oscillatory process variables is then revealed by Granger causality test and is visualized in the form of causality diagram. The simplification of causal connectivity in the diagram is performed according to the understanding of process knowledge and the final simplest causality diagram, which represents the main oscillation propagation paths, is achieved by the automated cutting-off thresh-old search, with which less significant causality pathways are filtered out. The source of the oscillation disturbance can be identified intuitively through the final causality diagram. Both simulated and real plant data tests are presented to demonstrate the effectiveness and feasibility of the proposed method.     

基于典型相关分析和数据自回归处理的BP神经网络在聚丙烯熔融指数预报中的应用

由于聚丙烯生产是一个大量参数相互耦合的强非线性过程,使得传统的机理建模受到一定的限制。提出基于典型相关分析和数据自回归处理的BP神经网络软测量建模,通过可测变量来推知聚丙烯熔融指数。应用典型相关分析选择与输出熔融指数关系较大的独立输入变量,数据自回归处理校正一系列带有误差的量测数据,而BP神经网络用来刻画过程的非线性特征。最后,将提出的算法应用到聚丙烯大型生产工艺中进行熔融指数的预报建模并进行实例仿真,仿真结果表明该算法有较强的建模精度。     

电厂过程监测与故障诊断的一种MSPC方法

讨论了大型电站具有随机时变特性过程的状态监测与故障诊断。基于多变量统计过程（MSPC）理论的主元素分析（PCA）技术,分辨故障事件的原因与结果。为适应随机时变特性的过程特点,将典型的PCA与改进的递推和快速滑动窗型PCA算法进行了比较。给出了递推快速滑动窗型算法（MWPCA）,通过与典型的PCA算法在故障诊断应用中的比较,表明了新算法对过程特性变化的自适应能力和计算效率。给出了在现代电站过程中应用FMWPCA算法将诊断与过程特性变量结合的图示方法应用实例。     

Melt index prediction by RBF neural network with an ICO‐VSA hybrid optimization algorithm

Melt index (MI) is considered as one of the most significant parameter to determine the quality and the grade of the practical polypropylene polymerization products. A novel ICO‐VSA‐RNN (RBF neural network with ICO‐VSA algorithm) MI prediction model is proposed based on radial basis function (RBF) neural network and improved chaos optimization (ICO), and variable‐scale analysis (VSA), where the ICO is first added and then combined with the VSA to overcome the defects of ICO and VSA, then the parameters of the RBF neural network are optimized with them. At last, the RBF neural network model for MI prediction model is developed. Further researches on the optimal RBF neural network model of MI prediction are carried out with the data from a real industrial plant, and the prediction results show that the performance of this prediction model is much better than the RBF neural network model without optimization. © 2012 Wiley Periodicals, Inc. J Appl Polym Sci, 2012     

一种基于新型蚁群算法的聚丙烯熔融指数预报模型

聚丙烯熔融指数的实时预报非常重要却十分困难,提出了一种经过新型蚁群算法优化后的PCA-RBF神经网络方法进行熔融指数预报。PCA将原始数据从高维空间映射到低维空间,剔除冗余信息和提取过程特征;RBF神经网络则用来拟合输入与输出之间的非线性关系;最后用适用于连续空间寻优问题的新型蚁群算法对RBF神经网络权值进行优化。实际生产数据的研究结果,表明了所提出的熔融指数预报模型的准确性和可靠性。     

Melt index prediction by RBF neural network optimized with an adaptive new ant colony optimization algorithm

Melt index (MI) is a crucial indicator in determining the product specifications and grades of polypropylene (PP). The prediction of MI, which is important in quality control of the PP polymerization process, is studied in this work. Based on RBF (radial basis function) neural network, a soft‐sensor model (RBF model) of the PP process is developed to infer the MI of PP from a bunch of process variables. Considering that the PP process is too complicated for the RBF neural network with a general set of parameters, a new ant colony optimization (ACO) algorithm, N‐ACO, and its adaptive version, A‐N‐ACO, which aim at continuous optimizing problems are proposed to optimize the structure parameters of the RBF neural network, respectively, and the structure‐best models, N‐ACO‐RBF model and A‐N‐ACO‐RBF model for the MI prediction of propylene polymerization process, are presented then. Based on the data from a real PP production plant, a detailed comparison research among the models is carried out. The research results confirm the prediction accuracy of the models and also prove the effectiveness of proposed N‐ACO and A‐N‐ACO optimization approaches in solving continuous optimizing problem. © 2010 Wiley Periodicals, Inc. J Appl Polym Sci, 2010     

递归核PCA及其在非线性过程自适应监控中的应用

PCA、PLS作为常用的多变量统计监控算法，一般适用于线性、定常的过程。针对实际工业过程的时变、非线性特性，提出了一种递归核PCA（RKPCA）方法用于非线性过程的自适应监控。RKPCA算法通过将递归奇异值分解推广到核空间，给出了核形式描述的递归KPCA算法，运算复杂度比KPCA明显降低，保证非线性监控模型能够在线更新。在Alstom工业燃气发生装置上的自适应监控表明，所提出的RKPCA算法能够及时跟踪非线性过程的时变特征，保证了监控模型的有效性。     

基于距离空间统计量分析的多模态过程无监督故障检测

工业过程往往运行于多个生产模态,针对多模态过程数据的空间分布特点,提出了一种新的基于样本距离空间统计量分析的故障检测方法(DSSA)。首先用每一个样本与其训练集样本中的邻居之间的k个最近邻距离之差来表示该样本,将样本从原始变量空间映射到对应的距离空间中。然后在距离空间中通过移动窗口的方式计算各阶统计量,最后对由各阶统计量组成的统计量样本进行主元分析(PCA)。将DSSA方法、PCA方法以及另一种基于k近邻规则的多模态故障检测方法(FD-kNN)应用于TE过程中,仿真结果表明DSSA方法对多模态故障检测更为有效。