International Journal of Industrial Engineering and Productional Research
Volume:19 Issue: 4, Dec 2008

Supplier selection is one of the most important activities of purchasing departments. This importance is increased even more by new strategies in a supply chain, because of the key role suppliers perform in terms of quality, costs and services which affect the outcome in the buyer’s company. Supplier selection is a multiple criteria decision making problem in which the objectives are not equally important. In practice, vagueness and imprecision of the goals, constraints and parameters in this problem make the decision making complicated. Simultaneously, in this model, vagueness of input data and varying importance of criteria are considered. In real cases, where Decision- Makers (DMs) face up to uncertain data and situations, the proposed model can help DMs to find out the appropriate ordering from each supplier, and allows purchasing manager(s) to manage supply chain performance on cost, quality, on time delivery, etc. An additive weighted model is presented for fuzzy multi objective supplier selection problem with fuzzy weights. The model is explained by an illustrative example.

In this paper, a two-echelon supply chain, which includes two products based on the following considerations, has been studied and the bullwhip effect is quantified. Providing a measure for bullwhip effect that enables us to analyze and reduce this phenomenon in supply chains with two products is the basic purpose of this paper. Demand of products is presented by the first order vector autoregressive time series and ordering system is established according to order up to policy. Moreover, lead-time demand forecasting is based on moving average method because this forecasting method is used widely in real world. Based on these assumptions, a general equation for bullwhip effect measure is derived and there is a discussion about non-existence of an explicit expression for bullwhip effect measure according to the present approach on the bullwhip effect measure. However, bullwhip effect equation is presented for some limited cases. Finally, bullwhip effect in a two-product supply chain is analyzed by a numerical example.

: In this paper, we present the stochastic version of Maximal Covering Location Problem which optimizes both location and allocation decisions, concurrently. It’s assumed that traveling time between customers and distribution centers (DCs) is uncertain and described by normal distribution function and if this time is less than coverage time, the customer can be allocated to DC. In classical models, traveling time between customers and facilities is assumed to be in a deterministic way and a customer is assumed to be covered completely if located within the critical coverage of the facility and not covered at all outside of the critical coverage. Indeed, solutions obtained are so sensitive to the determined traveling time. Therefore, we consider covering or not covering for customers in a probabilistic way and not certain which yields more flexibility and practicability for results and model. Considering this assumption, we maximize the total expected demand which is covered. To solve such a stochastic nonlinear model efficiently, simulation and genetic algorithm are integrated to produce a hybrid intelligent algorithm. Finally, some numerical examples are presented to illustrate the effectiveness of the proposed algorithm.

The advent of e-commerce has prompted many manufacturers to redesign their traditional channel structure by engaging in direct sales. In this paper, we present a dual channel inventory model based on queuing theory in a manufacturer-retailer supply chain, consisting of a traditional retail channel and a direct channel which stocks are kept in both upper and lower echelon. The system receives stochastic demand from the both channel which each channel has an independent demand arrival rate. A lost-sales model which no backorder is allowed is supposed. The replenishment lead times are assumed independent exponential random variables for both warehouse and the retail store. Under the replenishment inventory policy, the inventory position is kept constant at a base-stock level. To analyze the chain performance, an objective function included holding and lost sales costs is defined. At the end, a proposed algorithm named, Best Neighborhood (BN) is used to find a good solution for inventory and the results are compared with Simulated Annealing (SA) solutions.

Although control charts are very common to monitoring process changes, they usually do not indicate the real time of the changes. Identifying the real time of the process changes is known as change-point estimation problem. There are a number of change point models in the literature; however most of the existing approaches are dedicated to normal processes. In this paper we propose a novel approach based on clustering techniques to estimate Shewhart control chart change-point when a sustained shift is occurrs in the process mean. For this purpose we devise a new clustering mechanism, a new similarity measure and a new objective function. The proposed approach is not only capable of detecting process change-points, but also automatically estimates the true values of the out-of-control parameters of the process. We also compare the performance of the proposed approach with existing methods.

In this paper, a single machine sequencing problem is considered in order to find the sequence of jobs minimizing the sum of the maximum earliness and tardiness with idle times (n/1/I/ETmax). Due to the time complexity function, this sequencing problem belongs to a class of NP-hard ones. Thus, a special design of a simulated annealing (SA) method is applied to solve such a hard problem. To compare the associated results, a branch-and-bound (B&B) method is designed and the upper/lower limits are also introduced in this method. To show the effectiveness of these methods, a number of different types of problems are generated and then solved. Based on the results of the test problems, the proposed SA has a small error, and computational time for achieving the best result is very small.

In this paper Semi-Markov models are used to forecast the triple dimensions of next earthquake occurrences. Each earthquake can be investigated in three dimensions including temporal, spatial and magnitude. Semi-Markov models can be used for earthquake forecasting in each arbitrary area and each area can be divided into several zones. In Semi-Markov models each zone can be considered as a state of proposed Semi-Markov model. At first proposed Semi-Markov model is explained to forecast the three mentioned dimensions of next earthquake occurrences. Next, a zoning method is introduced and several algorithms for the validation of the proposed method are also described to obtain the errors of this method.