JOURNAL OF OPTIMIZATION IN INDUSTRIAL ENGINEERING (JOURNAL OF INDUSTRIAL ENGINEERING)
Year:2012 | Volume:6 | Issue:1 (10)|Papers Count:8

In this paper, the problem of determining the quality level, lead time for order delivery and price of a product produced by a manufacturer is considered. In this problem the demand for the product is influenced by all three decision variables: price, lead time and quality level. To formulate the demand function, a fuzzy rule base that estimates the demand value based on the three decision variables is developed. To do so, the linguistic knowledge of experts in the form of if-then rules is used to establish the fuzzy system. Moreover, in order to solve the problem, a genetic algorithm integrating the fuzzy rule base is proposed. Finally, to support the validity of the proposed solution, a numerical study is provided.

Estimation of the conceptual costs in construction projects can be regarded as an important issue in feasibility studies. This estimation has a major impact on the success of construction projects. Indeed, this estimation supports the required information that can be employed in cost management and budgeting of these projects. The purpose of this paper is to introduce an intelligent model to improve the conceptual cost accuracy during the early phases of the life cycle of projects in construction industry. A computationally efficient model, namely support vector machine model, is developed to estimate the conceptual costs of construction projects. The proposed neural network model is trained by a cross validation technique in order to produce the reliable estimations. To demonstrate the performance of the proposed model, two powerful intelligent techniques, namely nonlinear regression and back-propagation neural networks (BPNNs), are provided. Their results are compared on the basis of the available dataset from the related literature in construction industry. The computational results illustrate that the presented intelligent model performs better than the other two powerful techniques.

This study aims to deal with a more realistic combined problem of project scheduling and material ordering. The goal is to minimize the total material holding and ordering costs by determining the starting time of activities along with material ordering schedules subject to some constraints. The problem is first mathematically modelled. Then a hybrid simulated annealing and genetic algorithm is proposed to solve it. In addition, some experiments are designed and the Taguchi method is employed to both tune the parameters of the proposed algorithm and to evaluate its performance. The results of the performance analysis show the efficiency of the proposed methodology.

The sensitivity analysis for multi-attribute decision making (MADM) problems is important for two reasons: First, the decision matrix as the source of the results of a decision problem is inaccurate because it sorts the alternatives in each criterion inaccurately. Second, the decision maker may change his opinions in a time period because of changes in the importance of the criteria and in the policy of the organization over time. This in turn makes problem solving really time-consuming. Therefore, the best solution is to do sensitivity analysis.In this regard, this paper considers a sensitivity analysis in the QUALIFLEX method which is a compromise ranking method used for multi-criteria decision making (MCDM).

Logistic network design is one of the most important strategic decisions in supply chain management that has recently attracted the attention of many researchers. Transportation network design is then one of the most important fields of logistic network. This study is concerned with designing a multi-stage and multi-product logistic network. At first, a mixed integer nonlinear programming model (MINLP) is formulated that minimizes transportation and holding costs. Then, a hybrid priority-based Genetic Algorithm (pb-GA) and simulated annealing algorithm (SA) is developed in two phases to find the optimal solution. The solution is represented by a matrix and a vector. Response Surface Methodology (RSM) is also used to adjust the significant parameters of the algorithm. Finally, several test problems are generated which show that the proposed metaheuristic algorithm can find good solutions in reasonable time spans.

In this study, we discuss the capacitated facility location-allocation problem with uncertain parameters in which the uncertainty is characterized by given finite numbers of scenarios. In this model, the objective function minimizes the total expected costs of transportation and opening facilities subject to the robustness constraint. To tackle the problem efficiently and effectively, an efficient hybrid solution algorithm based on several meta-heuristics and an exact algorithm is put forward. This algorithm generates neighborhoods by combining the main concepts of variable neighborhood search, simulated annealing, and tabu search and finds the local optima by using an algorithm that uses an exact method in its framework. Finally, to test the algorithms’ performance, we apply numerical experiments on both randomly generated and standard test problems. Computational experiments show that our algorithm is more effective and efficient in term of CPU time and solutions quality in comparison with CPLEX solver.

In this paper we consider the non-preemptive variant of a multi-mode resource constrained project scheduling problem (MRCPSP) with mode identity, in which a set of project activities is partitioned into disjoint subsets while all activities forming one subset have to be processed in the same mode. We present a depth-first branch and bound algorithm for the resource constrained project scheduling problem with mode identity. The proposed algorithm is extended with some bounding rules to reduce the size of branch and bound tree. Finally, some test problems are solved and their computational results are reported.

This paper studies the hybrid flow shop scheduling where the optimization criterion is the minimization of total tardiness. First, the problem is formulated as a mixed integer linear programming model. Then, to solve large problem sizes, an artificial immune algorithm hybridized with a simple local search in form of simulated annealing is proposed. Two experiments are carried out to evaluate the model and the algorithm. In the first one, the general performance of the model and the proposed algorithm is experimented. In the next one, the presented algorithm is compared against some other algorithms. The results support high performance of the proposed algorithm.