A flexible ANN-GA-multivariate algorithm for assessment and optimization of machinery productivity in complex production units

This paper presents a flexible algorithm based on artificial neural networks (ANNs), genetic algorithms (GAs), and multivariate analysis for performance assessment and optimization of complex production units (CPUs) with respect to machinery productivity indicators (MPIs). Multivariate techniques include data envelopment analysis (DEA), principal component analysis (PCA) and numerical taxonomy (NT). Two case studies are considered to show the applicability of the proposed approach. In the first case, the machinery productivity indicators are categorized into four standard classes as availability, machinery stoppage, random failure and value added and production value. In the second case, the productivity of production units in terms of health, safety, environment and ergonomics indicators is evaluated. The flexible algorithm is capable of handling both linearity and complexity of data sets. Moreover, ANN and GA are efficiently applied to cover nonlinearity and complexity of CPUs. The results are also validated and verified by the internal mechanism of the algorithm. The algorithm is applied to a large set of production units to show its superiority and applicability over conventional approaches. Results show that, in the case of having non-linear data sets, ANN outperforms GA and conventional approaches. The flexible algorithm of this study may be easily extended to other units for assessment and optimization of CPUs with respect to machinery indicators.     

Entropic component analysis and its application in geological data

We present an entropic component analysis for identifying key parameters or variables and the joint effects of various parameters that characterize complex systems. This approach identifies key parameters through solving the variable selection problem. It consists of two steps. First, a Bayesian approach is utilized to convert the variable selection problem into the model selection problem. Second, the model selection is achieved uniquely by evaluating the information difference of models by relative entropies of these models and a reference model. We study a geological sample classification problem, where a brine sample from Texas and Oklahoma oil field is considered, to illustrate and examine the proposed approach. The results are consistent with qualitative analysis of the lithology and quantitative discriminant function analysis. Furthermore, the proposed approach reveals the joint effects of the parameters, while it is unclear from the discriminant function analysis. The proposed approach could be thus promising to various geological data analysis.     

Designing classifier fusion systems by genetic algorithms

We suggest two simple ways to use a genetic algorithm (GA) to design a multiple-classifier system. The first GA version selects disjoint feature subsets to be used by the individual classifiers, whereas the second version selects (possibly) overlapping feature subsets, and also the types of the individual classifiers. The two GAs have been tested with four real data sets: heart, Satimage, letters, and forensic glasses. We used three-classifier systems and basic types of individual classifiers (the linear and quadratic discriminant classifiers and the logistic classifier). The multiple-classifier systems designed with the two GAs were compared against classifiers using: all features; the best feature subset found by the sequential backward selection method; and the best feature subset found by a CA. The GA design can be made less prone to overtraining by including penalty terms in the fitness function accounting for the number of features used.     

Stepwise variable selection and maximum likelihood estimation of smoothing factors of kernel functions for nonparametric discriminant functions evaluated by different criteria

Kernel functions are used to estimate the probability density functions of variables for nonparametric discriminant analysis. In connection with stepwise variable identification a stepwise maximum likelihood estimation procedure for the estimation of smoothing factors of the kernel functions is developed. This procedure allows a step-by-step estimation of smoothing factors for every variable which is considered to be added to the model or which is examined to substitute a variable in a model. Different criteria for model evaluation in stepwise discriminant analysis are discussed. Beside criteria, like distance and dependence functions and the error and nonerror rate, a criterion which considers the ratio of probability densities of different classes at point x is proposed for stepwise variable identification. An application of the procedures described in this study to a medical decision problem shows the importance of stepwise parameter estimation of kernel functions for nonparametric discriminant analysis and the role of different model evaluation criteria for the selection of the best subset of variables.     

Hybrid genetic algorithms for feature selection

This paper proposes a novel hybrid genetic algorithm for feature selection. Local search operations are devised and embedded in hybrid GAs to fine-tune the search. The operations are parameterized in terms of their fine-tuning power, and their effectiveness and timing requirements are analyzed and compared. The hybridization technique produces two desirable effects: a significant improvement in the final performance and the acquisition of subset-size control. The hybrid GAs showed better convergence properties compared to the classical GAs. A method of performing rigorous timing analysis was developed, in order to compare the timing requirement of the conventional and the proposed algorithms. Experiments performed with various standard data sets revealed that the proposed hybrid GA is superior to both a simple GA and sequential search algorithms.     

基于T-PLS贡献图方法的故障诊断技术

多变量统计过程监控对于复杂工业过程是一种有效的故障检测和诊断技术. 最小二乘(或称潜空间投影)模型是多变量统计过程监控中常用的一种投影模型, 能够同时对过程数据和质量数据进行建模. 讨论了一种新的基于全潜空间投影模型的故障诊断技术. 全潜空间投影模型中有4个检测统计量. 提出了一种新的T2贡献图计算方法, 对于所有检测统计量, 得到了相应的贡献图算法. 为了确定一个变量是否发生了故障, 计算所有变量贡献图的控制限. 该技术可以将辨识到的故障变量分为与Y有关和与Y无关的两类. 基于Tennessee Eastman过程的案例研究表明了该技术的有效性.     

局部Fisher判别分析算法及仿真研究*

复杂化工过程常被多种类型的故障损坏,正常的训练数据无法建立准确的操作模型。为了提高复杂化工过程中故障的检测和分类能力,传统无监督Fisher判别分析(Fisher Discriminant Analysis,FDA)算法无法在多模态故障数据中的应用,本文提出基于局部Fisher判别分析(Local Fisher Discriminant Analysis,LFDA)的故障诊断方法。首先计算训练数据的局部类内和类间离散度矩阵,寻找LFDA的投影方向;其次把训练数据和测试数据向投影向量上投影,提取特征向量;最后计算特征向量间的欧氏距离,运用KNN分类器进行分类。把提出的LFDA方法应用到Tennessee Eastman(TE)过程,监控结果表明,LFDA的效果好于FDA和核Fisher判别分析(Kernel Fisher Discriminant Analysis,KFDA),说明LFDA方法在分类及检测不同类的故障方面具有高准确性及高灵敏度的优势。     

一种基于数据重构的KPCA故障识别方法

针对核主元分析(KPCA)方法只能实现故障检测,但无法实现故障变量识别的问题,提出一种基于数据重构的KPCA故障变量识别方法。采用改进的数据重构方法对各参数进行重构,然后利用故障识别指数对监控参数进行故障变量识别。通过对某型涡扇发动机进行实验的结果表明,该方法能够准确识别故障变量,从而有助于维护人员分析故障原因,初步确定可能的故障源,大大缩短故障定位及排故的时间,可预防重大事故的发生。     

Differentiable coarse graining

Coarse graining is defined in terms of a commutative diagram. Necessary and sufficient conditions are given in the continuously differentiable case. The theory is applied to linear coarse grainings arising from partitioning the population space of a simple Genetic Algorithm (GA). Cases considered include proportional selection, binary tournament selection, ranking selection, and mutation. A nonlinear coarse graining for ranking selection is also presented. A number of results concerning “form invariance” are given. Within the context of GAs, the primary contribution made is the illustration of a technique by which coarse grainings may be analyzed. It is applied to obtain a number of new coarse graining results.     

Combining genetic algorithms with BESO for topology optimization

This paper proposes a new algorithm for topology optimization by combining the features of genetic algorithms (GAs) and bi-directional evolutionary structural optimization (BESO). An efficient treatment of individuals and population for finite element models is presented which is different from traditional GAs application in structural design. GAs operators of crossover and mutation suitable for topology optimization problems are developed. The effects of various parameters used in the proposed GA on the optimization speed and performance are examined. Several 2D and 3D examples of compliance minimization problems are provided to demonstrate the efficiency of the proposed new approach and its capability of obtaining convergent solutions. Wherever possible, the numerical results of the proposed algorithm are compared with the solutions of other GA methods and the SIMP method.     

基于混合准则的软测量建模辅助变量选择方法

软传感器在工业中被广泛应用于预测与产品质量密切相关的关键过程变量,这些变量很难在线测量。要建立一个高精度的软传感器,选择合适的辅助变量是至关重要的。针对这个问题,本文通过耦合训练集的BIC准则以及验证集的MSE准则得到一个混合整数非线性规划问题,并将该MINLP问题分成内外两层结构,外层采用遗传算法对二元整数变量进行寻优,内层在整数变量固定之后退化成了较易于求解的非线性规划问题。在此基础上经过进一步分析提出了基于混合准则的变量选择方法,然后将所得辅助变量子集代入BP神经网络进行软测量建模。最后,通过4组案例对所提出方法进行验证。结果表明,所提出方法建立的软测量模型具有较好的预测性能。     

Allocating data to distributed-memory multiprocessors by genetic algorithms

We present three genetic algorithms (GAs) for allocating irregular data sets to multiprocessors. These are a sequential hybrid GA, a coarse-grain GA and a fine-grain GA. The last two are based on models of natural evolution that are suitable for parallel implementation; they have been implemented on a hypercube and a Connection Machine. Experimental results show that the three GAs evolve good suboptimal solutions which are better than those produced by other methods. The GAs are also robust and do not show a bias towards particular problem configurations. The two parallel GAs have reasonable execution times, with the coarse-grain GA producing better solutions for the allocation of loosely synchronous computations.     

逐步贝叶斯判别分析中的变量优化方法研究

判别分析中特征变量是影响判别结果的决定性因素,选取适当的特征变量组合可以提高正判率、减少计算量。介绍了贝叶斯判别和逐步判别法的基本原理,分析了目前出现的一些特征变量优化方法,以油气解释评价中的贝叶斯判别应用为例,对于逐步贝叶斯判别中的变量优化方法进行了研究和总结,提出了变量的多步优化策略和分步多模型优化策略,包含了从变量范围选择、数据预处理、特征变量提取到初步筛选和逐步判别的完整过程,使得正判率不断优化,最终得到了较为满意的判别结果。     

A new fault diagnosis method of multimode processes using Bayesian inference based Gaussian mixture contribution decomposition

This paper presents a novel Bayesian inference based Gaussian mixture contribution (BIGMC) method to isolate and diagnose the faulty variables in chemical processes with multiple operating modes. The statistical confidence intervals of traditional principal component analysis (PCA) based T² and SPE diagnostics rely upon the assumption that the operating data follow a multivariate Gaussian distribution approximately and therefore may not be able to determine the faulty variables in multimode non-Gaussian processes accurately. As an alternative solution, the proposed BIGMC method first identifies the multiple Gaussian modes corresponding to different operating conditions and then integrates the Mahalanobis distance based variable contributions across all the Gaussian clusters through Bayesian inference strategy. The derived BIGMC index is of probabilistic feature and includes all operation scenarios with posterior probabilities as weighting factors. The Tennessee Eastman process (TEP) is used to demonstrate the utility of the proposed BIGMC method for fault diagnosis of multimode processes. The comparison of the single-PCA and multi-PCA based contribution approaches shows that the BIGMC method can effectively identify the leading faulty variables with superior diagnosis capability.     

子空间半监督Fisher判别分析

Fisher判别分析寻找一个使样本数据类间散度与样本数据类内散度比值最大的子空间, 是一种很流行的监督式特征降维方法. 标注样本数据所属的类别通常需要大量的人工, 消耗大量的时间, 付出昂贵的成本. 为了解决同时利用有类别信息的样本数据和没有类别信息的样本数据用于寻找降维子空间的问题, 我们提出了一种子空间半监督Fisher判别分析方法. 子空间半监督Fisher判别分析寻找这样一个子空间, 这个子空间即保留了从有类别信息的样本数据中学习的类别判别结构, 也保留了从有类别信息的样本数据和没有类别信息的样本数据中学习的样本结构信息. 我们还推导了基于核的子空间半监督Fisher判别分析方法. 通过人脸识别实验验证了本文算法的有效性.     

Keep–Best Reproduction: A Local Family Competition Selection Strategy and the Environment it Flourishes in

This paper presents a comparison of two genetic algorithms (GAs) for constrained ordering problems. The first GA uses the standard selection strategy of roulette wheel selection and generational replacement (STDS), while the second GA uses an intermediate selection strategy in addition to STDS. This intermediate selection strategy keeps only the superior offspring and replaces the inferior offspring with the superior parent. We call this selection strategy Keep–Best Reproduction (KBR). The effect of recombination alone, mutation alone and both together are studied. We compare the performance of the different selection strategies and discuss the environment that each selection strategy needs to flourish in. Overall, KBR is found to be the selection strategy of choice. We also present empirical evidence that suggests that KBR is more robust than STDS with regard to operator probabilities and works well with smaller population sizes.     

Online monitoring and fault identification of mean shifts in bivariate processes using decision tree learning techniques

With modern data collection system and computers used for on-line process monitoring and fault identification in manufacturing processes, it is common to monitor more than one correlated process variables simultaneously. The main problems in most multivariate control charts (e.g., T ² charts, MCUSUM charts, MEWMA charts) are that they cannot give direct information on which variable or subset of variables caused the out-of-control signals. A Decision Tree (DT) learning based model for bivariate process mean shift monitoring and fault identification is proposed in this paper under the assumption of constant variance-covariance matrix. Two DT classifiers based on the C5.0 algorithm are built, one for process monitoring and the other for fault identification. Simulation results show that the proposed model can not only detect the mean shifts but also give information on the variable or subset of variables that cause the out-of-control signals and its/their deviate directions. Finally a bivariate process example is presented and compared with the results of an existing model.     

The theoretical analysis of FDA and applications

Representation and embedding are usually the two necessary phases in designing a classifier. Fisher discriminant analysis (FDA) is regarded as seeking a direction for which the projected samples are well separated. In this paper, we analyze FDA in terms of representation and embedding. The main contribution is that we prove that the general framework of FDA is based on the simplest and most intuitive FDA with zero within-class variance and therefore the mechanism of FDA is clearly illustrated. Based on our analysis, ε-insensitive SVM regression can be viewed as a soft FDA with ε-insensitive within-class variance and L₁ norm penalty. To verify this viewpoint, several real classification experiments are conducted to demonstrate that the performance of the regression-based classification technique is comparable to regular FDA and SVM.     

Mapping natural habitats using remote sensing and sparse partial least square discriminant analysis

This work presents a novel approach for mapping the spatial distribution of natural habitats in the ‘Foothills of Larzac’ Natura 2000 listed site located in a French Mediterranean biogeographical region. Sparse partial least square discriminant analysis was used to analyse two RapidEye data sets (June 2009 and July 2010) with the purpose of choosing the most informative spectral, textural, and thematic variables that allow discrimination of habitat classes. The sparse partial least square discriminant analysis selected relevant and stable variables for the discrimination of habitat classes that could be linked to ecological or biophysical characteristics. It also gave insight into the similarities and differences between habitat classes with comparable physiognomic characteristics. The highest user accuracy was obtained for dry improved grasslands (u = 91.97%) followed by riparian ash woods (u = 88.38%). These results are very encouraging given that these two classes were identified in Annex 1 of the EC Habitats Directive as of Community interest. Due to limited data input requirements and its computational efficiency, the approach developed in this article is a good alternative to other types of variable selection approaches in a supervised classification framework and can be easily transferred to other Natura 2000 sites.