A nonlinear quality-related fault detection approach based on modified kernel partial least squares

In this paper, a new nonlinear quality-related fault detection method is proposed based on kernel partial least squares (KPLS) model. To deal with the nonlinear characteristics among process variables, the proposed method maps these original variables into feature space in which the linear relationship between kernel matrix and output matrix is realized by means of KPLS. Then the kernel matrix is decomposed into two orthogonal parts by singular value decomposition (SVD) and the statistics for each part are determined appropriately for the purpose of quality-related fault detection. Compared with relevant existing nonlinear approaches, the proposed method has the advantages of simple diagnosis logic and stable performance. A widely used literature example and an industrial process are used for the performance evaluation for the proposed method.     

基于分块多向主成分分析的翻车机液压系统故障诊断

采用基于物理模型和统计模型的方法对翻车机液压系统故障进行监测。根据翻车机液压系统各阶段的工作机理将监测变量分块,使每一块中的变量间因果关系更加明确,再对各块分别建立多向主成分分析(MPCA)监测模型。将分块MPCA模型和全变量MPCA模型应用于压车缸泄漏故障的监测,结果表明分块MPCA模型对微小泄漏更加敏感,具有较高的故障识别率。     

基于变量分块的KDLV-DWSVDD间歇过程故障检测算法研究

非线性动态间歇过程中,测量变量存在不同的序列相关性,且变量间的交叉相关性会体现在不同的采样时刻上,然而传统检测方法没有考虑这种变量间的相关性,通常将所有变量视为独立或相关关系进行特征提取,不能充分提取到故障信息的特征,造成监测效果不佳.因此,提出一种基于变量分块的核动态潜变量-动态加权支持向量数据描述(KDLV-DWS...     

Fluid dynamic mechanisms of particle flow causing ductile and brittle erosion

Traditional prediction of erosion focuses on the use of velocity and impact angle of particles as independent variables in analytically derived models. This approach is most suitable for numerical predictions of erosion in disperse flow fields where particle trajectories may easily be followed prior to impact. For dense particle flows, the prediction of individual particle or particle cluster movement is nearly never attempted by following trajectories. Instead, two-fluid Eulerian–Eulerian approaches are used in which a continuous particle fluid phase is considered.The present study shows that the impact velocity and angle of attack of particles at the eroding surface are difficult to obtain for dense flows, thus being difficult to consider as parameters for predicting erosion. Instead, it is proposed that the normal and the shearing components of the viscous dissipation of the particulate phase are more suitable as independent flow variables governing the erosion process. These variables describe deformation and cutting wear processes, respectively, and are readily derived from the flow field.Eulerian erosion models are proposed, based on these independent variables. It is possible to implement previous results and theories concerning the material–mechanical interaction between the abrasive and an eroding surface to achieve model improvements. In this work, only a simple model taking into account a threshold elastic strain limit is proposed, to more correctly model the deformation wear.The particle-flow boundary condition — a partial-slip condition — significantly influences the erosion process, particularly the cutting erosion. The boundary condition depends on parameters such as the local particle phase flow, the mean diameter and the sharpness of the abrasive as well as the surface roughness.A simple 2D test application — a jet stream of particles impinging a tilted plate — is presented, and the qualitative angular behaviour of ductile and brittle erosion is reproduced at the target position. A scheme is presented for determination of material constants and suitable boundary conditions to be used in the proposed erosion models.     

装备试验质量成本管理浅析

装备试验质量成本管理是装备试验质量管理的一个重要方面，研究成本管理对于完善质量管理具有重要的意义。结合装备试验的特点，在对质量成本的构成进行分析的基础上，探讨了如何对成本进行管理，分析了成本的核算、分析、控制等问题。     

Performance monitoring method based on balanced partial least square and Statistics Pattern Analysis

For process monitoring, it's of significance to pay more attention to some performance indexes related to quality, economy or security. To this end, a novel performance monitoring method based on the prediction of the performance indexes is proposed in this paper to promote the efficiency of process monitoring. In this study, firstly, the process variables are classified into two categories according to the correlation with the performance indexes. Based on the two categories of variables, a balanced Partial Least Square algorithm is proposed by constructing an enhanced objective function to predict the performance indexes which cannot be measured online. Then, the prediction residual is modeled for monitoring via the Statistics Pattern Analysis to capture the variation in the performance indexes. Finally, two Simulink examples and a practical example are utilized for illustration and validation.     

基于模糊层次分析法的工程机械质量评价

工程机械从设计、装配到使用有一个实现过程,每一个环节都包含了多种与机械产品的质量密切相关的评价指标,比较分析这些评价指标,有利于决策者把握重点,提高产品的可靠性和经济性,结合工程机械特点,建立影响机械质量的指标体系,综合运用模糊数学和层次分析结构,建立模糊层次分析模型,提出了隶属度和权重系数的分配方法,通过对某型工程机械的质量评价,对评价方法进行详细介绍,评价结果表明,该方法科学合理,具有较高的实用性和通用性.     

Human critical success factors for kaizen and its impacts in industrial performance

This article presents the results of a survey with 13 main activities related to human factors that are executed during kaizen implementation process and were integrated in four independent latent variables (management commitment, education, communication and motivation) that are associated to 14 benefits obtained after its implementation that were grouped in three dependent latent variables (process, workers and customers). The survey was applied to persons with responsibilities in continuous improvement programs and projects in companies located in Mexico. Independent and dependent variables were integrated in a structural equation model that was evaluated using partial least squares algorithms WarpPLS® for finding causal relations among them. Results indicate that management commitment and education are the main factors that guarantee the success for kaizen implementation programs, but that is moderated by a good communication for having good operational process performance for better workers and customer satisfaction.     

A robust hybrid model integrating enhanced inputs based extreme learning machine with PLSR (PLSR-EIELM) and its application to intelligent measurement

In this paper, a robust hybrid model integrating an enhanced inputs based extreme learning machine with the partial least square regression (PLSR-EIELM) was proposed. The proposed PLSR-EIELM model can overcome two main flaws in the extreme learning machine (ELM), i.e. the intractable problem in determining the optimal number of the hidden layer neurons and the over-fitting phenomenon. First, a traditional extreme learning machine (ELM) is selected. Second, a method of randomly assigning is applied to the weights between the input layer and the hidden layer, and then the nonlinear transformation for independent variables can be obtained from the output of the hidden layer neurons. Especially, the original input variables are regarded as enhanced inputs; then the enhanced inputs and the nonlinear transformed variables are tied together as the whole independent variables. In this way, the PLSR can be carried out to identify the PLS components not only from the nonlinear transformed variables but also from the original input variables, which can remove the correlation among the whole independent variables and the expected outputs. Finally, the optimal relationship model of the whole independent variables with the expected outputs can be achieved by using PLSR. Thus, the PLSR-EIELM model is developed. Then the PLSR-EIELM model served as an intelligent measurement tool for the key variables of the Purified Terephthalic Acid (PTA) process and the High Density Polyethylene (HDPE) process. The experimental results show that the predictive accuracy of PLSR-EIELM is stable, which indicate that PLSR-EIELM has good robust character. Moreover, compared with ELM, PLSR, hierarchical ELM (HELM), and PLSR-ELM, PLSR-EIELM can achieve much smaller predicted relative errors in these two applications.     

A multi-objective optimization method based on Gaussian process simultaneous modeling for quality control in sheet metal forming

A new modeling method, related to multiple inputs and multiple outputs (MIMO), simultaneously based on Gaussian process (GP), is proposed to optimize the combinations of process parameters and improve the quality control for multi-objective optimization problems in sheet metal forming. In the MIMO surrogate model, for the use of the system information in processing and the accuracy of the model, quantitative and categorical input variables are both taken into account in GP simultaneously. Firstly, a general method is proposed for constructing covariance functions for GP simultaneous MIMO surrogate model based on correlation matrices. These covariance functions must be able to incorporate the valid definitions of both the spatial correlation based on quantitative input variables and the cross-correlation based on categorical input variables. Secondly, the unrestrictive correlation matrices are constructed by the hypersphere decomposition parameterization, thus directly solving optimization problems with positive definite constraints is needless, and the computational complexity is simplified. Compared with independent modeling method, the proposed GP simultaneous MIMO model has higher accuracy and needs less number of estimated parameters. Moreover, the cross-correlation between the outputs (quality indexes) obtained by proposed model provides some reference to further develop quality intelligent control strategies. Finally, a drawing-forming process of auto rear axle housing is taken as an example to validate the proposed method. The results show that the proposed method can effectively decrease the crack and wrinkle in sheet metal forming.     

A study on an axial-type 2-D turbine blade shape for reducing the blade profile loss

Losses on the turbine consist of the mechanical loss, tip clearance loss, secondary flow loss and blade profile loss etc.,. More than 60 % of total losses on the turbine is generated by the two latter loss mechanisms. These losses are directly related with the reduction of turbine efficiency. In order to provide a new design methodology for reducing losses and increasing turbine efficiency, a two-dimensional axial-type turbine blade shape is modified by the optimization process with two-dimensional compressible flow analysis codes, which are validated by the experimental results on the VKI turbine blade. A turbine blade profile is selected at the mean radius of turbine rotor using on a heavy duty gas turbine, and optimized at the operating condition. Shape parameters, which are employed to change the blade shape, are applied as design variables in the optimization process. Aerodynamic, mechanical and geometric constraints are imposed to ensure that the optimized profile meets all engineering restrict conditions. The objective function is the pitchwise area averaged total pressure at the 30 % axial chord downstream from the trailing edge. 13 design variables are chosen for blade shape modification. A 10.8 % reduction of total pressure loss on the turbine rotor is achieved by this process, which is same as a more than 1 % total-to-total efficiency increase. The computed results are compared with those using 11 design variables, and show that optimized results depend heavily on the accuracy of blade design.     

Prediction of Surface Topography at the End of Sliding Running-In Wear Based on Areal Surface Parameters

Running in is a complex process, and it significantly influences the performance and service life of wear components as the initial phase of the entire wear process. Surface topography is an important feature of wear components. Therefore, it is reasonable to investigate the running-in process with the help of surface topography for improvement. Because the surface roughness after running in is independent of the nature of initial roughness, it is difficult to predict the surface topography after running in based on unworn surface topography. Aiming to build a connection of surface topographies before and after the running-in process, a black-box model predicting surface topography after the running-in process was established based on least-squares support vector machine (LS-SVM), and the areal surface evaluation parameters were adopted as model variables. To increase the adaptability of the predictive model, the main factors of the work condition were also taken into consideration. The prediction effect and sensitivity of the model were tested and analyzed. The analysis indicates that the hybrid property of surface topographies before and after running in is closely related. Moreover, the surface topography after running in is influenced more by the initial surface topography than by the work condition.     

Monitoring multivariate–attribute processes based on transformation techniques

Control charts are widely used in monitoring the quality of a product or a process. In most of the cases, quality of a product or a process can be characterized by two or more correlated quality characteristics. Many control charts have been proposed for monitoring multivariate or multi-attribute quality characteristics, separately, but sometimes the correlated variables and attribute quality characteristics represents the quality of a process. In this paper, the use of four transformation methods is proposed to monitor the multivariate–attribute processes. In the first one, the distribution of correlated variables and attribute quality characteristics are transformed to approximate multivariate normal distribution, and then the transformed data are monitored by multivariate control charts including T ² and MEWMA. Based on the second transformation method, the correlated variables and attribute quality characteristics are transformed, such that the correlation between the quality characteristics approaches to zero, then univariate control charts are used in monitoring the transformed data. In the third and fourth proposed methods, a combination of two transformation methods is used to make the quality characteristics independent and to transform them to normal distribution. The difference between the third and fourth method is the order of using the transformation techniques. The performance of the proposed methods is evaluated by using simulation studies in terms of average run length criterion. Finally, the proposed approach is applied to a real dataset.     

Number of distinct data categories and gage repeatability and reproducibility. A double (but single) requirement

Measurement systems analysis has been widely used in many industries, especially those related to the automotive sector as part of statistical process control practices. Over the years, new elements have been added to the initial requirements for system acceptance, centered on the values of repeatability and reproducibility (integrated in indexes as GRR and GRR%). One of these elements, the number of distinct data categories (NDC) has been treated as an independent requirement. In this paper, we show that NDC and GRR are not independent and that they are in fact both linked by an exact relationship. This therefore makes the double requirement redundant, confusing and unnecessary, except maybe to enhance some conceptual questions.     

Tolerance chart balancing with a complete inspection plan taking account of manufacturing and quality costs

This paper introduces a mathematical model of tolerance chart balancing for machining process planning under complete inspection. The criteria considered in this study are based on the combined effects of manufacturing cost and quality loss, under constraints such as process capability limits, product design specifications, and product quality requirements. Manufacturing costs include the machining cost, part cost, inspection cost, reworking cost, and replacement cost. The machining cost is expressed in geometrical decreasing functions, which represent tolerances to be assigned. Process variability is expressed in quadratic loss functions, which represent the deviation between the part measurement and the target value. An example is presented to demonstrate the proposed model. A comparison made with previous methods shows that the proposed model minimizes the total cost of manufacturing activities and quality-related issues in machining process planning, particularly in the early stages. Moreover, the applications are not be limited to machining process planning but can also be used in other forms of production planning.     

Convolution and correlation: a case study of scanning imaging and analysis systems

The relationship between convolution/correlation operation and the data acquisition process of the scanning microscope and spectrometer families is analyzed. It is shown that a coordinate or event sensitive detector, and the intrinsic or extrinsic property of the specimen response, are two important factors in distinguishing the data acquisition mode of such systems. Four types of convolution- and correlation-based modes are extracted to illuminate the physical characteristics of scanning imaging and analysis systems by focusing on the probe, specimen, detector, and their relationships. Criteria for identifying these modes are explored. In addition, the physical meanings of general existing coefficients between the independent variables of convolution and correlation are investigated.     

浅谈焊接机器人系统自主集成实施的过程控制

随着人力劳动成本的迅速上升,工业焊接机器人在我国汽车焊装工艺中使用的频率越来越高.以SGMW自主集成焊接机器人系统为例,从项目管理的角度进行了探讨.对焊接机器人及其控制系统前期设计、安装调试及其带来的各种意义做出了相关介绍,供相关人士参考.     

Simultaneous process mean and process tolerance determination with adjustment and compensation for precision manufacturing process

Conventional process planning of manufacturing operations presets fixed process means and process tolerances for all operations and allows outputs to be distributed around these fixed values, as long as the final outputs meet acceptable specifications. Most of these approaches consider process means and process tolerances to be independent decision variables in process planning with the resultant process means equal or close to the design targets of the blueprint dimensions. Furthermore, these approaches assume that process variability is small in comparison to the quality requirement, and that the phenomena of process shifting or deterioration are not factors of manufacturing operations. For these reasons, conventional approaches to process planning are inappropriate for high value, and precision manufacturing process, particularly of a complex part. Hence, this study introduces a process optimization model which considers process means and process tolerances simultaneously, with sequential operation adjustment to reduce process variability, and with part compensation to offset process shifting.