基于KVAE-OCCA的质量相关故障检测方法及应用

为了解决非线性过程质量相关故障检测问题,提出了一种名为关键变量自编码器-正交典型相关分析(KVAE-OCCA)的方法.首先,为了挑选出与质量变量具有相关性的过程变量,计算过程变量和质量变量的互信息,选择具有较大互信息的过程变量.然后,利用自编码器对选择出的过程变量进行无监督学习,实现特征提取和降维.其次,利用正交典型相...     

基于多组典型相关变量的因果关系发现算法

现有的因果关系发现算法主要基于单个观察变量本身之间的因果关系,无法适用于多组观察变量,为此提出了一种多组典型相关变量的因果关系发现算法。首先,引入多组典型相关变量建立多组典型相关变量的线性非高斯无环模型并提出对应的目标函数;然后,采用梯度上升的方法求解目标函数,构建多组典型相关变量的因果关系网络。模拟实验验证了该算法的有效性,并在移动基站数据上发现了一批有价值的多组无线网络性能指标间的因果关系。     

Reduced order optimization for model predictive control using principal control moves

In order to reduce the computational complexity of model predictive control (MPC) a proper input signal parametrization is proposed in this paper which significantly reduces the number of decision variables. This parametrization can be based on either measured data from closed-loop operation or simulation data. The snapshots of representative time domain data for all manipulated variables are projected on an orthonormal basis by a Karhunen-Loeve transformation. These significant features (termed principal control moves, PCM) can be reduced utilizing an analytic criterion for performance degradation. Furthermore, a stability analysis of the proposed method is given. Considerations on the identification of the PCM are made and another criterion is given for a sufficient selection of PCM. It is shown by an example of an industrial drying process that a strong reduction in the order of the optimization is possible while retaining a high performance level.     

Visual‐interactive Exploration of Interesting Multivariate Relations in Mixed Research Data Sets

The analysis of research data plays a key role in data‐driven areas of science. Varieties of mixed research data sets exist and scientists aim to derive or validate hypotheses to find undiscovered knowledge. Many analysis techniques identify relations of an entire dataset only. This may level the characteristic behavior of different subgroups in the data. Like automatic subspace clustering, we aim at identifying interesting subgroups and attribute sets. We present a visual‐interactive system that supports scientists to explore interesting relations between aggregated bins of multivariate attributes in mixed data sets. The abstraction of data to bins enables the application of statistical dependency tests as the measure of interestingness. An overview matrix view shows all attributes, ranked with respect to the interestingness of bins. Complementary, a node‐link view reveals multivariate bin relations by positioning dependent bins close to each other. The system supports information drill‐down based on both expert knowledge and algorithmic support. Finally, visual‐interactive subset clustering assigns multivariate bin relations to groups. A list‐based cluster result representation enables the scientist to communicate multivariate findings at a glance. We demonstrate the applicability of the system with two case studies from the earth observation domain and the prostate cancer research domain. In both cases, the system enabled us to identify the most interesting multivariate bin relations, to validate already published results, and, moreover, to discover unexpected relations.     

带拒绝箱覆盖问题的局内算法

作为对装箱覆盖问题的推广,提出带拒绝的装箱覆盖问题.设有许多等长的一维箱子,给定一个物品集,每个物品有两个参数:长度和费用.物品可以放入箱子也可被拒绝放入箱子,每个物品只准放入一只箱子中,每只箱子中的物品容量总和至少为箱子容量,一旦箱子中的物品长度达到要求则需启用新箱.如果物品被放入箱中,则产生费用.该问题是一个新的组合优化问题,在内部互联网信息管理等问题中有着广泛的应用背景.给出一个求解该问题的局内近似算法C-FF,分析其最坏情况渐近性能比为1/2,并给出了相应的实验结果.     

基于核典型相关性-熵成分分析的工业过程质量监测方法

工业过程多变量、数据高维度和非线性的特点使得对其质量监测及质量相关的故障诊断变得复杂.融合核熵成分分析(KECA)及典型相关分析(CCA)方法的思想,进行特征提取降维的同时确保所提取特征与质量变量的最大相关性,提出一种新的质量相关的工业过程故障检测方法.首先,采用KECA对输入数据进行核空间的映射及特征提取,同时融合CCA算法思想使得所提取特征与质量变量间关联最大化;然后,构建监测统计量并用Parzen窗估计其控制限,用于过程的故障检测;最后,运用所提方法对带钢热连轧工业过程实际生产数据进行分析,并与其他4种传统非线性算法对比分析,实验结果验证了所提方法的准确性、有效性及先进性.     

Blind Source Separation Exploiting Higher-Order Frequency Dependencies

Blind source separation (BSS) is a challenging problem in real-world environments where sources are time delayed and convolved. The problem becomes more difficult in very reverberant conditions, with an increasing number of sources, and geometric configurations of the sources such that finding directionality is not sufficient for source separation. In this paper, we propose a new algorithm that exploits higher order frequency dependencies of source signals in order to separate them when they are mixed. In the frequency domain, this formulation assumes that dependencies exist between frequency bins instead of defining independence for each frequency bin. In this manner, we can avoid the well-known frequency permutation problem. To derive the learning algorithm, we define a cost function, which is an extension of mutual information between multivariate random variables. By introducing a source prior that models the inherent frequency dependencies, we obtain a simple form of a multivariate score function. In experiments, we generate simulated data with various kinds of sources in various environments. We evaluate the performances and compare it with other well-known algorithms. The results show the proposed algorithm outperforms the others in most cases. The algorithm is also able to accurately recover six sources with six microphones. In this case, we can obtain about 16-dB signal-to-interference ratio (SIR) improvement. Similar performance is observed in real conference room recordings with three human speakers reading sentences and one loudspeaker playing music     

Monitoring of quality-relevant and quality-irrelevant blocks with characteristic-similar variables based on self-organizing map and kernel approaches

Variables in quality-related process monitoring can be divided into quality-relevant and quality-irrelevant groups depending on the correlation with the quality indicator. These variables can also be separated into multiple sets in which variables are closely relevant to one another because of the interdependence of the process. Block monitoring with reasonable variable partition and reliable model can distinguish quality-related and quality-unrelated faults and improve monitoring performance. A block monitoring method based on self-organizing map (SOM) and kernel approaches is proposed. After collecting and normalizing the sample data including process variables and quality ones, the data matrix is transposed. The inverted samples are used as the input of SOM, and variables with the same behavioral characteristic and a close correlation are topologically mapped in a similar area. Accordingly, samples can be visually blocked into quality-relevant and independent subspaces. Given the nonlinearity of industrial process, kernel partial least squares (KPLS) and kernel principal component analysis (KPCA) are employed to monitor the two types of blocks. The information provided by fault detection can reveal the effects on quality indicators and the location of faults. Finally, the effectiveness of SOM-KPLS/KPCA is evaluated using a numerical example and the Tennessee–Eastman process.     

Performance and precision of histogram calculation on GPUs: Cosmological analysis as a case study

Histogram calculation is an essential part of many scientific analyses. In Cosmology, histograms are employed intensively in the computation of correlation functions of galaxies, as part of Large Scale Structure studies. Among the most commonly used ones are the two-point, three-point and the shear–shear correlation functions. In these computations, the precision of the calculation of the counts in each bin is a key element for achieving the highest accuracy. In order to accelerate the analysis of increasingly larger datasets, GPU computing is becoming widely employed in this field. However, the recommended histogram calculation procedure becomes less precise when bins become highly populated in these sort of algorithms. In this work, an alternative implementation to correct this problem is proposed and tested. This approach is based on distributing the creation of histograms between the CPU and GPU. The implementation is tested using three cosmological analyses with observational data. The results show an increased performance in terms of accuracy while keeping the same execution time.     

一种监督的局部保持典型相关分析算法

从模式分类的角度出发,提出一种监督的局部保持典型相关分析(SLPCCA),通过最大类内成对样本与其近邻间的权重相关性,因而能有效利用样本类别信息的同时保持数据的局部流形结构,并且融合判别型典型相关分析(DCCA)的鉴别信息而不受总类别数的限制。此外,为了提取数据的非线性特征,在核方法的基础上又提出一种核化的SLPCCA(KSLPCCA)。在ORL、Yale、AR和FERET等人脸数据库的实验结果表明,该算法比其他传统的典型相关分析方法具有更好的识别效果。     

基于PLS交叉积矩阵非相似度分析的MPC性能监控与诊断

针对传统基于输出协方差矩阵的性能监控方法未充分考虑过程变量与输出变量之间的相关性问题,提出一种基于偏最小二乘（Partial least squares,PLS）交叉积矩阵非相似度分析的性能监控与诊断方法,用于多变量模型预测控制（Model predictive control,MPC）系统.首先,考虑模型预测控制系统的控制结构,构造包含预测误差的增广过程变量与输出变量相关性的PLS交叉积矩阵,通过非相似度分析方法将交叉积矩阵的非相似度比较转化为转换矩阵特征值的比较.然后提取转换矩阵中表征最大非相似度的l个特征值构造实时性能指标,对MPC系统进行性能监控.检测到性能下降后,进一步利用转换矩阵的特征值诊断性能恶化源.Wood-Berry二元精馏塔上的仿真结果表明,所提方法能够有效地提高监控性能,并准确地定位性能恶化源.     

Enhanced canonical variate analysis with slow feature for dynamic process status analytics

Process dynamics is widely presented in industrial processes, which can be perceived as temporal correlations. Negligence of dynamic information may result in misleading monitoring results. Therefore, explicit exploration of dynamic information is crucial to process monitoring. In this paper, a new data-driven algorithm called enhanced canonical variate analysis with slow feature (ECVAS) and corresponding monitoring strategy are proposed for dynamic process monitoring. First, a new objective function is defined with two goals, which attempts to extract slowly varying latent variables in addition to high temporal correlation. Hence, the latent variables called slow canonical variables (SCVs) would capture valuable dynamic information and be isolated from static information and fast-varying noises. Second, the process dynamics has been explored in detail by concurrently monitoring of temporal correlations and varying speed. Therefore, the proposed method achieves in-depth understanding of process dynamics under control actions and helps identify normal changes in operating conditions. Third, process static information and dynamic information have been separately monitored, contributing to a fine-scale identification of process variations. Finally, the validity of the proposed strategy is illustrated with an industrial scale multiphase flow experimental rig and a real thermal power process.     

Quantization-based clustering algorithm

In this paper, a quantization-based clustering algorithm (QBCA) is proposed to cluster a large number of data points efficiently. Unlike previous clustering algorithms, QBCA places more emphasis on the computation time of the algorithm. Specifically, QBCA first assigns the data points to a set of histogram bins by a quantization function. Then, it determines the initial centers of the clusters according to this point distribution. Finally, QBCA performs clustering at the histogram bin level, rather than the data point level. We also propose two approaches to improve the performance of QBCA further: (i) a shrinking process is performed on the histogram bins to reduce the number of distance computations and (ii) a hierarchical structure is constructed to perform efficient indexing on the histogram bins. Finally, we analyze the performance of QBCA theoretically and experimentally and show that the approach: (1) can be easily implemented, (2) identifies the clusters effectively and (3) outperforms most of the current state-of-the-art clustering approaches in terms of efficiency.     

Dimension reduction via principal variables

For many large-scale datasets it is necessary to reduce dimensionality to the point where further exploration and analysis can take place. Principal variables are a subset of the original variables and preserve, to some extent, the structure and information carried by the original variables. Dimension reduction using principal variables is considered and a novel algorithm for determining such principal variables is proposed. This method is tested and compared with 11 other variable selection methods from the literature in a simulation study and is shown to be highly effective. Extensions to this procedure are also developed, including a method to determine longitudinal principal variables for repeated measures data, and a technique for incorporating utilities in order to modify the selection process. The method is further illustrated with real datasets, including some larger UK data relating to patient outcome after total knee replacement.     

一种新的有监督的局部保持典型相关分析算法

从模式识别的角度出发,在局部保持典型相关分析的基础上,提出一种有监督的局部保持典型相关分析算法(SALPCCA)。该方法在构造样本近邻图时将样本的类别信息考虑在内,由样本间的距离度量确定权重,建立样本间的多重权重相关,通过使同类内的成对样本及其近邻间的权重相关性最大,从而能够在利用样本的类别信息的同时,也能保持数据的局部结构信息。此外,为了能够更好地提取样本的非线性信息,将特征集映射到核特征空间,又提出一种核化的SALPCCA(KSALPCCA)算法。在ORL、Yale、AR等人脸数据库上的实验结果表明,该方法较其他的传统典型相关分析方法有着更好的识别效果。     

有序判别典型相关分析

多视图学习方法通过视图间互补信息的融合,达到增强单一视图方法的鲁棒性并提升学习性能的目的.典型相关分析(canonical correlation analysis,简称CCA)是一种重要的多视图信息融合技术.其研究的是针对同一组目标两组不同观测数据间的相关性,目标是得到一组相关性最大的投影向量.但当面对标号有序的分类任务时,CCA因没有利用类信息和类间有序信息,造成了对分类性能的制约.为此,通过将有序类信息嵌入CCA进行扩展,发展出有序判别典型相关分析(ordinal discriminative canonical correlation analysis,简称OR-DisCCA).实验结果表明, OR-DisCCA的性能比相关方法更优.     

一种新的多约束尺寸可变的装箱问题

多约束尺寸可变的装箱问题作为经典装箱问题的扩展,具有极为广泛的应用背景。在以货车运输为主的物流公司的装载环节中,运输成本不仅仅由车厢的空间利用率决定。分析了该类装箱问题与传统的集装箱装载问题的区别,并据此给出了一种新的尺寸可变装箱问题的定义。除了经典装箱问题中物品体积这一参数,还引入了物品类型、箱子类型等参数,建立了数学模型,将经典的FFD（First Fit Decreasing）算法进行了推广,提出了新的算法MFFD,并分析了相关的算法复杂性。最后对FF、FFD以及MFFD算法进行了模拟实验,实验结果表明,在相关参数符合均匀分布的条件下,MFFD算法效果较好。     

Transforming input variables for RBFN based on PCA-ASH multivariate correlation analysis and its application

The mutual information (MI) based on averaged shifted histogram (ASH) probability density estimator is considered as a good indicator of relevance between input variables and output variable. However, it cannot deal with redundant input variables problem. Therefore, a method integrates principal component analysis (PCA) with MI is proposed for radial basis function network (RBFN) to improve the predicting performance of RBFN. Firstly, PCA is employed to characterize the PCs from original variables, among which there is non-correlation. Secondly, MI based on ASH is applied to select the several closest correlation PCs with output variable as the new input variables. Finally, PCA-ASH-RBFN is employed to develop the housing price model based on the Boston housing data set. The result shows that PCA-ASH-RBFN has better prediction and robust performance than PCA-RBFN and RBFN integrating with robust feature selection for input variables.

     

New heuristics for one-dimensional bin-packing

Several new heuristics for solving the one-dimensional bin packing problem are presented. Some of these are based on the minimal bin slack (MBS) heuristic of Gupta and Ho. A different algorithm is one based on the variable neighbourhood search metaheuristic. The most effective algorithm turned out to be one based on running one of the former to provide an initial solution for the latter. When tested on 1370 benchmark test problem instances from two sources, this last hybrid algorithm proved capable of achieving the optimal solution for 1329, and could find for 4 instances solutions better than the best known. This is remarkable performance when set against other methods, both heuristic and optimum seeking.Scope and purposePacking items into boxes or bins is a task that occurs frequently in distribution and production. A large variety of different packing problems can be distinguished, depending on the size and shape of the items, as well as on the form and capacity of the bins (H. Dyckhoff and U. Finke, Cutting and Packing in Production and Distribution: a Typology and Bibliography, Springer, Berlin, 1992). Similar problems occur in minimising material wastage while cutting pieces into particular smaller ones and in the scheduling of identical processors in order to minimise total completion time. This work addresses the basic packing problem, known as the one-dimensional bin packing problem, where it is required to pack a number of items into the smallest possible number of bins of pre-specified equal capacity. Even though this problem is simple to state, it is NP hard, i.e., it is unlikely that there exists an algorithm that could solve every instance of it in polynomial time. Solution of more general realistic packing problems is probably contingent upon the availability of effective and computationally efficient solution procedures for the basic problem. In this work we present several heuristics capable of doing that. Extensive computational testing attests to the power of these heuristics, as well as to their computational efficiency.     

Semi-supervised PLVR models for process monitoring with unequal sample sizes of process variables and quality variables

As the key indicators of chemical processes, the quality variables, unlike process variables, are often difficult to obtain at the high frequency. Obtaining the data of quality variables is expensive, so the data are only collected as a small portion of the whole dataset. It is common to see in both continuous and batch processes that the sample sizes of process variables and quality variables are unequal. To effectively integrate two different observation sources, including quality variables collected at a low frequency and process variables sampled at a high rate, a semi-supervised probabilistic latent variable regression model (SSPLVR) is proposed in this article. It enhances the performance monitoring of the variations of process variables and quality variables. The proposed semi-supervised model is applied to continuous and batch processes respectively. The SSPLVR model calibrated by the expectation-maximization algorithm is derived and the corresponding statistics is also systematically developed for the fault detection. Finally, two simulated case studies, TE benchmark for a continuous process problem and the penicillin fermentation for a batch process problem, are presented to illustrate the effectiveness of the proposed method.