AMIRKABIR
Year:2003 | Volume:14 | Issue:55-D|Papers Count:6

This paper focuses on the appropriate selection of suppliers and assigning order quantities to them. This problem contains multiple objectives and their related constraints and utilizes fuzzy decision making in its solution approach. In the proposed model, goals, constraints, variables and coefficients are all fuzzy. It is shown that with the application of the fuzzy methodology, the multiy objective problem is converted to a single one that can be solved easily.

Trace element geochemistry, mineral paragenesis, and fluid inclusion studies, have identified two stages of hydrothermal alteration and related mineralization within the Sungun skarn. The first stage (stage I) was produced by the early pulses of Fe-rich magmatically derived fluids, which is coincident with potassic alteration at the main stock, and caused the formation of oscillatory zoned garnets, followed by pyroxene (diopside and hedenbergite) + wollastonite + magnetite and rare Mo at high temperatures (~450 oC). The second stage (II) occurred at lower temperature (~360 oC) when the system collapsea to cool down, and is characterised by tremolite-actinolite, epidote, quartz, calcite, apatite, titanite, hematite crystallization and copper deposition. It is concluded that the Cu mineralization was due to decreas off 01 and increase of pH in lower temperature range (400 DC) in last stage of alteration in the Sungun skarn.In addition, the detailed mineralogical studies present that in the Sungun skarns, the garnet displays complex oscillatory zoning due to fluctuations in the Al+3/Fe+3 ratio and varies in composition from andradite 100 to andradite78-grossular22 Individual zones are typically composed either of near end-member andradite or andradite containing approximately 20 mole % grossular. The oscillation from andradite to less grossular-bearing zones is interpreted to reflect fluid and rock-dominated control of Fe+3and Al+3 activity, respectively, and batch-rise or episodic influx of fluid. Introduction of fresh batches of Fe-rich fluid caused crystallization of near end-member andradite garnet since the system was buffered by fluid.

Determination of optimal repair policy of a component (or machine) is one of the main subjects at maintenance planning. In this field the main approach is to use a finite state approximation Markov Decision Process (MDP). This technique requires a known long time for its as computation. In this paper an alternative approach is introduced. Assuming a stochastically repair cost, the optimum solution is shown to satisfy a first order differential equation. Also an asymptotic formula for optimal repair policy is derived. Numerical solutions are obtained for some weibull and gamma distributed time to falure distributions. This new method is faster and easier than previous methods.

Comparison matrix formed by pairwise comparisons is the main part of Analytic Hierarchy Process (AHP) method. Since the values in the matrix are given by a Decision Maker (DM), there is a potential context for incocsistency. In fuzzy AHP, where terms or numbers are used, the inconsistency is a crucial issue. To make an appropriate decision, the inconsistency in decision deal with inconsistency in fuzzy comparison matrix. In the first model, DM's judgments are modified where the direction of judgments are remained unchanged. The second model allows the reversing the direction of judgments. The proposed models aim to minimize the number of changes in earlier DM's judgments. Two examples solved by the proposed models, are included for illustrative purposes.

Assembly line balancing problems are important problems in production management systems. Based on the nature of systems the jobs processing time may be exact, stochastic, or fuzzy. In this article a genetic algorithm approach for both type of fuzzy assembly line balancing problems is introduced. The reported results and comparisons show that proposed algorithm is promising.

Various type of uncertainties and imprecision are inherent in real inventory problem. they are classically modeled using the approaches from the probability theory. However there are uncertainties that cannot be appropriately treated by usual probabilistic models. In this paper, a model for inventory control in fuzzy sense is presented. In this model the parameters of holding cost, ordering cost and the price of each item are considered as normal fuzzy trapezoidal number and the parameter of demand (consumption) is considered as a crisp number. To obtain the membership functions, in addition to using the "a-cut" method, a method based on the concept of "extension" principle is devised and applied.