T. N. Goh

A modified Weibull extension with bathtub-shaped failure rate function

Abstract Models with bathtub-shaped failure rate function are useful in reliability analysis, and particularly in reliability related decision making and cost analysis. The traditional Weibull distribution is, however, unable to model the complete lifetime of systems with a bathtub-shaped failure rate function. In this paper, a new model, which is useful for modeling this type of failure rate function, is presented. The model can also be seen as a generalization of the Weibull distribution. Parameter estimation methods are studied for this new distribution. Examples and results of comparison are shown to illustrate the applicability of this new model.

A comparative study of neural network and Box-Jenkins ARIMA modeling in time series prediction

This paper aims to investigate suitable time series models for repairable system failure analysis. A comparative study of the Box-Jenkins autoregressive integrated moving average (ARIMA) models and the artificial neural network models in predicting failures are carried out. The neural network architectures evaluated are the multi-layer feed-forward network and the recurrent network. Simulation results on a set of compressor failures showed that in modeling the stochastic nature of reliability data, both the ARIMA and the recurrent neural network (RNN) models outperform the feed-forward model; in terms of lower predictive errors and higher percentage of correct reversal detection. However, both models perform better with short term forecasting. The effect of varying the damped feedback weights in the recurrent net is also investigated and it was found that RNN at the optimal weighting factor gives satisfactory performances compared to the ARIMA model.

Some effective control chart procedures for reliability monitoring

Abstract Control charts are widely used for process monitoring in the manufacturing industry. Little research is available on their use to monitor the failure process of components or systems, which is important for equipment performance monitoring. Some Shewhart control charts, especially those for the number of defects, can be used for monitoring the number of failures per fixed interval; however, they are not effective especially when the failure frequency becomes small. A recent control scheme based on the cumulative quantity between observations of defects has been proposed which can be easily adopted to monitor the failure process for exponentially distributed inter-failure time. An investigation of its use for reliability monitoring is presented in this paper and the scheme can be easily extended to monitor inter-failure times that follow other distributions such as the Weibull distribution. Furthermore, the scheme is extended to the monitoring of time required to observe a fixed number of failures. The advantages of this scheme include the fact that the scheme does not require any subjective sample size, can be used for both high and low reliability items and can detect process improvement even in a high-reliability environment.

A study of project selection and feature weighting for analogy based software cost estimation

A number of software cost estimation methods have been presented in literature over the past decades. Analogy based estimation (ABE), which is essentially a case based reasoning (CBR) approach, is one of the most popular techniques. In order to improve the performance of ABE, many previous studies proposed effective approaches to optimize the weights of the project features (feature weighting) in its similarity function. However, ABE is still criticized for the low prediction accuracy, the large memory requirement, and the expensive computation cost. To alleviate these drawbacks, in this paper we propose the project selection technique for ABE (PSABE) which reduces the whole project base into a small subset that consist only of representative projects. Moreover, PSABE is combined with the feature weighting to form FWPSABE for a further improvement of ABE. The proposed methods are validated on four datasets (two real-world sets and two artificial sets) and compared with conventional ABE, feature weighted ABE (FWABE), and machine learning methods. The promising results indicate that project selection technique could significantly improve analogy based models for software cost estimation.

Cumulative quantity control charts for monitoring production processes

Two commonly used statistical quality control charts, the c-chart and u-chart, are unsatisfactory for monitoring high-yield processes with low defect rates. To overcome this difficulty, a new type of control chart called the cumulative quantity control chart (CQC-chart) is introduced in this paper. The CQC-chart can be used no matter whether the process defect rate is low or not, and when the process defect rate is low or moderate the CQC-chart does not have the shortcoming of the c- and u-charts of showing up false alarm signals too frequently. The CQC-chart does not require rational subgrouping of samples (which is necessary for the c- and u-charts), and is appropriate for monitoring automated manufacturing processes.

CONTROL CHART FOR MULTIVARIATE ATTRIBUTE PROCESSES

Many industrial processes are multivariate in nature since the quality of a product depends on more than one variable. Multivariate control procedures can be used to capture the relationship between the variables and to provide more sensitive control than that provided by the application of univariate control procedures on each variable. Much has been done on the multivariate variable processes, such as embodied in control procedures based on Hotellings T 2 statistic. However, little work has been done to deal with the control of multivariate attribute processes, which is very important in practical production processes. In this paper, we develop a Shewhart-type control chart to deal with multivariate attribute processes, which is called the multivariate np chart (MNP chart). The control chart uses the weighted sum of the counts of nonconforming units with respect to all the quality characteristics as the plotted statistics. It enhances the efficiency of identifying the critical assignable cause when an ...

Statistical Models and Control Charts for High-Quality Processes

Preface. Acknowledgments. 1. Introduction. 2. Control Charts with Probability Limits. 3. Cumulative Count of Conforming (CCC) chart. 4. Process Improvement Detection. 5. Modified Implementation of Geometric Chart. 6. Some Extensions to the Geometric Model. 7. CUSUM and EWMA Procedures. 8. Monitoring of Multiple Process Characteristics. 9. Economic Design of Geometric Chart. 10. Monitoring and Adjustment of Trended Processes. References. Subject Index.

A systematic comparison of metamodeling techniques for simulation optimization in Decision Support Systems

Simulation is a widely applied tool to study and evaluate complex systems. Due to the stochastic and complex nature of real world systems, simulation models for these systems are often difficult to build and time consuming to run. Metamodels are mathematical approximations of simulation models, and have been frequently used to reduce the computational burden associated with running such simulation models. In this paper, we propose to incorporate metamodels into Decision Support Systems to improve its efficiency and enable larger and more complex models to be effectively analyzed with Decision Support Systems. To evaluate the different metamodel types, a systematic comparison is first conducted to analyze the strengths and weaknesses of five popular metamodeling techniques (Artificial Neural Network, Radial Basis Function, Support Vector Regression, Kriging, and Multivariate Adaptive Regression Splines) for stochastic simulation problems. The results show that Support Vector Regression achieves the best performance in terms of accuracy and robustness. We further propose a general optimization framework GA-META, which integrates metamodels into the Genetic Algorithm, to improve the efficiency and reliability of the decision making process. This approach is illustrated with a job shop design problem. The results indicate that GA-Support Vector Regression achieves the best solution among the metamodels.

A study of the connectionist models for software reliability prediction

Abstract When analysing software failure data, many software reliability models are available and in particular, nonhomogeneous Poisson process (NHPP) models are commonly used. However, difficulties posed by the assumptions, their validity, and relevance of these assumptions to the real testing environment have limited their usefulness. The connectionist approach using neural network models are more flexible and with less restrictive assumptions. This model-free technique requires only the failure history as inputs and then develops its own internal model of failure process. Their ability to model nonlinear patterns and learn from the data makes it a valuable alternative methodology for characterising the failure process. In this paper, a modified Elman recurrent neural network in modeling and predicting software failures is investigated. The effects of different feedback weights in the proposed model are also studied. A comparative study between the proposed recurrent architecture, with the more popular feedforward neural network, the Jordan recurrent model, and some traditional parametric software reliability growth models are carried out.

A statistical methodology for the analysis of the life-cycle of reusable containers

Abstract The effectiveness of recycling reusable containers as a means of spreading container costs depends on a knowledge of container life-cycle characteristics expressed through parameters such as trip-page, trip duration, loss rate, and expected useful life. Most existing estimation methods in industry are based on static ratios of inventory records. This paper describes, with numerical examples, a rigorous approach for data analysis and modeling which yields the needed parameters through a systematic examination of inherent statistical properties in routinely available data on container issues and returns. The results are important to functions such as pricing, marketing, accounting, and inventory and financial control in a container filling plant.