فصلنامه مطالعات مدیریت صنعتی
پیاپی 27 (زمستان 1391)

One of the trench that is used in quality desiccation is Quality Function Deployment. One of the factors which has been used in this method is accomplishing the house of quality and as a result is determining the degree of importance of the customer's requirements. Kano model is a kind of two dimensional model which help the producers to divide the consumer's requirements in to five categories including, basic, performance, attractive, indifferent quality and reverse quality. In the survey by identifying the customer's requirements we determine kind of them and then according to the improvement ratio and the degree of importance is assigned. The degrees of importance can be used for completing the house of quality which is one of the useful tools in QFD. This research is public algorithms that it can use for every product. For example it has been used in Cement Company of larestan, product II Portland cement.

Estimating the optimal number of clusters in an unsupervised partitioning of data sets has been a challenging area in recent years. These indices usually use two criteria called compactness and separation to evaluate the efficiency of the performed clustering. In this paper a new separation measure for ECAS cluster validity index, proposed by Fazel et al. [1] is identified, which uses Jaccard distance in order to consider the whole shape of clusters. Jaccard distance uses the size of intersection and union of fuzzy sets, giving the cluster validity index more information about the overlap and separation of clusters. This property results in high robustness of the proposed index dealing with various degrees of fuzziness in comparison with ECAS. To test the efficiency of the proposed index in comparison with nine other indices existing in the literature, 15 data sets (3 existing datasets and 12 artificial data sets) have been used. Computational results indicate robustness and high capability of the proposed index in comparison with previous indices.

In this research, we use Artificial Bee Colony (ABC) algorithm to solve cell to switch assignment problem (CTSAP) that is NP-hard. In CTSAP, there are cells and switches in which cells locations are predetermined. The objective of problem is optimal assigning of cells to switches with minimum cost. Here, we have two kinds of costs, handoff and cabling costs. Call handling capacity for every switches are given and equal. The model of our work is single homed that is each cell must connect to only one switch. The mathematical model is binary and nonlinear. The program is coded by MATLAB 7.8.0 (R2009a). After estimating parameters values of model, approving performance accuracy of code and adjusting control parameters, the efficiency of algorithm by determining experimental problems compared to Ant Colony Optimization (ACO) that is one of the best for solving this problem. Results show satisfactory performance of ABC algorithm.

Despite the extensive use of failure mode and effects analysis (FMEA) in the manufacturing and services, emphasis on the results of this method with concern to the disadvantages such as providing definitive results is not correct. This paper tries to clarify the main disadvantages of FMEA conventional method and explain the concept of cluster analysis and specifications of different models of C-Means to classify harmful modes and effects in the steel making factory of Iran's alloy steel company using FMEA and fuzzy clustering techniques. In this context, homogeneous harmful modes and effects based on occurrence, severity and frequency indicators with Fuzzy C-Means clustering (FCM) were identified. In this study, the fuzzy cluster analysis in terms of before and after improvement was performed by Data Engine 4.0 software. The results of this research will be useful for researchers, managers, and safety professionals for developing practical strategies in order to prevent the consequences of harmful modes and effects.

The Analytic Hierarchical Process (AHP) determines the relative importance of a set of alternatives in a multi-criteria decision problem. AHP has a hierarchial structure and based on pairwise comparisons of the project alternatives as well as pairwise comparisons of the multi criteria. Two separate algorithms have been presented for measuring the inconsistency ratioes of pairwise comparisons matrix and hierarchical structure. In this paper these algorithms and one of the approximation methods, have been extended for measuring the inconsistency ratioes and alternatives’ weights in fuzzy AHP. A numerical example is used to illustrate these methods.

In this paper, to make flexible job shop scheduling problem (FJSP) more realistic, an operational factor is considered in its model. This factor, which is called optimizing total consumed electric power per month, is known as the most important factor in calculation of the electric cost of the industries. Considering this factor, specifically after subsides elimination of the country, has became more important. In addition to this objective, two other common objectives, called complementation time and critical work load of machines, are considered. To solve the multi-objective model, two algorithms, called multi-objective biogeography-based optimization algorithm (MOBBO) and multi-objective harmony search algorithm (MOHS), are developed and introduced to scheduling area for the first time. Finally, by developing some famous libraries of the problem, performance of the algorithms is compared statistically.

The earned value management is a powerful and important technique in analyzing and controlling the project performance. While it allows exact measurement of project progress, it lets corrective actions in a timely manner. In fact, the earned value allows project managers to find out any project time and cost deviations by calculating the performance indices. In this paper, to improve the applicability of the traditional earned value technique, we develop an integrated approach by combining statistical quality control charts with traditional earned value technique, to monitor and control project time and cost performances. The results applied to a real construction project compete favorly against traditional approaches.