نشریه بین المللی مهندسی صنایع و مدیریت تولید
سال بیستم شماره 2 (تابستان 1388)

In this paper a multi-period inventory control problem will be analyzed. Periodic replenishment will happen totally stochastic and also the periods between two replenishments are independent and identically distributed random variables. indeed producer is encountered with customer in stochastic time. Also the decision variable has been chosen as an integer. In first model shortage will be back ordered and in second model shortage will be lost sale. Finally in third model combination of them will be considered. Three main specifications of the model which has led to be new are the stochastic period length, being multi-product and multi-constraint and the fact that the decision variables are integer. As this condition is deployed simultaneously therefore the created model is different from other models in the periodic review literature. The model of this problem is an integer nonlinear programming. Also for solving the model, the Simulated Annealing is used and for analyzing the output of the algorithms a numerical example will be exhibited.

In this paper the problem of job shop scheduling with parallel machines in each stages is discussed. The objective is to minimize the maximum completion time (makespan).�This problem is a combination of two classic problems of job shop and parallel machines which in this case parallel machines has been used as kind of flexibility in the job shop problem. The review of literature has shown that this problem has not been discussed yet. After presenting the mathematical mode, heuristic algorithms are used for solving this NP-hard problem. Regarding this, five algorithms are presented and a lower bound is developed. Finally all these algorithms have been analyzed. According to results the proposed algorithm of H2 works better than the others when there are few jobs. As the number of jobs increases H1 is more efficient than H2 asymptotically. Also the efficiency of H3 algorithm is the worst among the rest.

Vehicle Routing Problem with Time Windows (VRPTW) is an NP-Complete Optimization Problem. Even finding an optimal solution for small size problems is too hard and time-consuming. The objective of VRPTW is to use a fleet of vehicles with specific capacity to serve a number of customers with dissimilar demands and time window constraints at minimum cost, without violating the capacity and time window constraints. This problem has been solved with a number of heuristic and meta-heuristic solution algorithms and optimal or near optimal solutions gained. In this paper, a modified Ant Colony algorithm is proposed. In this algorithm we tried to simplify the solution procedure and computational complexities of ant colony meta-heuristic. To gain this capability, we sacrificed some computational accuracy. Testing the solution procedure on the Solomon test-problems showed that this algorithm is capable of generating relatively good solutions.

Considering customer’s perspective to evaluate Project management success beside time, cost and quality has gained more attention by researchers since the late 90s. Therefore a questionnaire based on potential factors from customer’s viewpoint for evaluating performance of project management in automotive parts industry was the main source of data collection in this research. Factor analysis was applied to identify underlying factors to develop an evaluation model. Afterwards the proposed model was applied to evaluate performance of project management in such projects. Results shows current weakness in managing risks and lack of awareness of its role in project management success is one of the main factors influencing performance of project management practices in these projects.

The group scheduling problem in the cellular manufacturing system is comprised of two levels of scheduling. At the first level, the sequence of parts in each part-family is determined, and then at the second level the sequence of part-families is determined. In this paper, the flow shop group scheduling is investigated in order to minimize the makespan. In traditional group scheduling problems, the parts processing times of are assumed to be constant and independent of the job sequence. This assumption is impractical in many situations because a worker’s ability and skill improve as a result of repeating the tasks, hence the part processing times decrease. This phenomenon is known as the learning effect.In this paper, a position-based learning model is used in the cellular manufacturing system in which each part processing time is dependent on its position in the parts sequence of part-family. The group scheduling problem is modeled with position-based learning effect and the part-families sequence-dependent setup times. Two different genetic algorithms and a heuristic method are developed to solve this problem and evaluated by the test problems. Numerical results obtained from evaluation of the proposed solution methods reveals that second genetic algorithm is better than other proposed methods regarding quality of the solutions.

The classical Job Shop Scheduling Problem (JSSP) is NP-hard problem in the strong sense. For this reason, different metaheuristic algorithms have been developed for solving the JSSP in recent years. The Particle Swarm Optimization (PSO), as a new metaheuristic algorithm, has applied to a few special classes of the problem. In this paper, a new PSO algorithm is developed for JSSP. First, a preference list of the generated solutions is prepared to obtain the feasibility. Then, a new method using factorial base numeral system, called factoradic approach, is developed to satisfy the validity of the generated solutions. This approach permits a one to one mapping of a solution in discrete space to a PSO particle position in continuous space. Since PSO is an evolutionary approach, some modifications are implemented to the algorithm. For examples, a simple greedy algorithm is developed to generate relatively good initial population or every solution obtained by PSO is also improved by a local search operator. To avoid trapping into local optima, a new velocity updating equation is considered. Furthermore, a Simulated Annealing (SA) approach is applied to the final solution obtained by PSO to improve it. Finally, the proposed hybrid algorithm is tested by some job shop benchmark problems. The results indicate the efficiency of the proposed hybrid PSO with respect to other algorithms exist in the literature for the considered problem.