نشریه بین المللی مهندسی صنایع و مدیریت تولید
سال بیست و چهارم شماره 4 (زمستان 1392)

Control charts are the most useful tools for controlling the processes statistically. The construction of the control charts requires the estimation of the process parameters using random sample data. Usually the classical estimators of the process parameters are used to construct the control charts. The classical estimators of the parameters of the processes generating autocorrelated data are sensitive to the presence of the outlier observations. Applying classical methods of estimation while outliers are present, introduce biased estimates of the model parameters which result in wrong interpretation of the control chart. In this research a method called Iteratively Robust Filtered Fast Tau (IRFFT) which is insensitive to the presence of the outliers is proposed for estimating the parameters of the autocorrelated models. The newly introduced estimators are used to construct robust control chart for autocorrelated data. The suggested robust control chart is compared with the control chart whose parameters are estimated using LS method. Results of the simulation study for the two methods indicate that the ARL for the suggested robust control chart is much smaller under different scenarios. The findings may be extended to the other time series models.

Control charts are powerful tools to monitor process. Design of control chart is usually performed through two ways, statistical and economical approaches Sometimes, a relationship, commonly called profile, between a response variable and one or more explanatory variables is defined for process monitoring. Various control charts proposed up to now to monitor profiles have not been designed economically. The importance of an economic design of profiles is significant. This is due to difference in some concepts and numerous of parameters. In this paper, for one of the monitoring approaches, which use three exponentially weighted moving average control charts simultaneously, cost function is presented and this approach is redesigned economically. In addition, the average run length calculating model is developed using the Markov chain method. This economic design is applied for a numerical example and solved by Nelder-Mead downhill simplex method and the optimal values of parameters are calculated.

In recent years, reliability as a design characteristic of products has widespread applications in many industries including defense and aerospace industries and became attractive issue. The main aim of reliability engineering is to improve system performance. Use of redundancy components is one of the ways for reliability improvement, although usually it is not possible to gain exact data in this type of systems to determine optimum number of redundancy components. Therefore, fuzzy logic has been introduced for assessing and analyzing the systems. This study presents a mathematical model for optimal allocation of the number of redundant components at each stage of multi-stage systems with structures such as parallel-series and standby to maximize system reliability subjecting to cost and weight constraints; which in this model reliability and cost and weight of each components and cost and weight of system are considered as fuzzy parameters. At last a numerical example regarding to triangular and trapezoidal fuzzy numbers and using different methods for defuzzification is surveyed and value of variables is computed with a simulated annealing meta-heuristic approach and the results are presented.

The main objective of performed researches in the field of multivariate statistical process control is to consider the correlation between multiple qualitative attributes for one step of process. In the second phase of the multivariate process control procedure, the rest of process is being studied whether it is under control, using the achieved control limits from the first phase and future observations. So, having found the outlier points of the first phase before the control limits to be computed, it is considered as an important issue. In order to detect these outlier points, Variety of techniques, that the majority of them rely on primary random samples, are proposed. These primary random points can effect on the precision of algorithms and final solution of the problem. In this paper, a robust estimator is issued applying hierarchical clustering technique that is not affected by outlier data in sample or unusual data, rather than the model assumptions and will detect the outlier points in multivariate control charts of the first phase in order to get them removed. Then, the proposed method is evaluated by creating the variety of scenarios from outlier points and the final outcome is compared with the Classical Hoteling and the least determinant covariance estimator. The evaluations represent that the proposed method detects more outlier points in less time rather than the former performed researches.

Camera is one of the most important types of modern communicational devices which failing on its operation may cause difficulties to achieve its goals. In order to protect camera lens from damages, especially in difficult missions at space, a type of cap called shutter can be put on the lens. Diagnosing all probable failures in such sensitive devices before testing can have a dramatic impact on modification of these systems. On the other hand, calculating of reliability in these systems due to lack of essential data in some cases subjected to some difficulties. Bayesian Network method is a powerful way in calculation of reliability for the mixed systems which can resolve the lack of data. In this paper firstly Functional Flow Block Diagram and the N2 matrix is drawn due to importance of sub-systems which are used in the shutter. Then the reliabilities of the components are computed through the methodology of Feasibility of Objects. In the next step, on the basis of shutter FTA; Bayesian Network is drawn. Finally the reliability of the shutter is calculated.

In this paper, the effect of estimation error on the performance of Sign non-parametric control chart is considered. A simulation program has been prepared that calculate ARLs and cumulative distribution functions in sign chart for Normal, Uniform, Laplace and Cauchy processes. The simulations results show effects of estimation error in heavy-tailed distributions is greater than light-tailed distributions. Also, the effect of sample size on the estimation error is considered. The results specify the heavy-tailed distributions need to more samples than light-tailed distributions.In this paper, the effect of estimation error on the performance of Sign non-parametric control chart is considered. A simulation program has been prepared that calculate ARLs and cumulative distribution functions in sign chart for Normal, Uniform, Laplace and Cauchy processes. The simulations results show effects of estimation error in heavy-tailed distributions is greater than light-tailed distributions. Also, the effect of sample size on the estimation error is considered. The results specify the heavy-tailed distributions need to more samples than light-tailed distributions.

Marketing campaigns has widely used warranty period as a quality indication of product offer. However this subject has been refered as a quality issue but the basic principle of such a period determination is a cost benefit trade off for every manufacturer.The model proposed in this paper is a renewable warranty policy in which the warranty period of components will extend after a failure. Furthermore an algorithm is presented that takes the reliability distribution of the components being analysed, in order to reach an optimal policy while considering a predetermined budget. Reliability determination and application of proposed algorithm has also been applied to powertrain system to represent the suggested algorithm.

Today, the flow shop environment is the most customary production area in industry. But, the need to increase the capacity or the capacity balance in different production processes has caused that in some working stations there exists more than one machine. This type of flow shops are usually named hybrid flow shop, multi processes flow shop, flexible flow shop or flow shop with parallel machines. In this article, the time scheduling of hybrid flow shop with the assumption of machine similarity is studied using a bi-criterion programming problem. Taking the fact that the proposed problem is an NP-hard type problem a genetic algorithm is developed and a sample problem is solved for illustration purposes.

The passenger line planning is a process of strategic long-term decision-making problem in the field of railway passenger planning. A line is a direct railway connection between two stations. A line is characterized by the origin and destination stations, the frequency per day, the route between these two stations, and the intermediate stops at passing railway stations. Various mathematical models have been proposed for passenger line planning problem, and different solution methods have been provided so far. Two solution methods based on column generation algorithm and genetic algorithm have been proposed in this article, the first alghorithm is defined in one master problem and two sub problems. Since the solution provided by column generation method is not of the integer number type, a heuristic algoadrithm has been proposed here for converting the results to integer numbers. The objective function for line planning problem in this article is to maximize the number of direct passengers. Direct passengers are the passengers that can travel from their origin station to their destination station without having to change trains. Results on the proposed solution methods, for twirty two test problems, are compared to those of solutions obtained via CPLEX software, one of the well-known solvers for solving both linear and mix integer programming problems using branch and bound algorithm. Results show that the proposed solution methods have high performance and accuracy.