Scientia Iranica
Volume:23 Issue: 1, 2016

This paper presents a set of multiple objective programming problems in a rough environment. These problems are classified into five classes according to the location of the roughness in the objective functions or the feasible set. We study the class in which all of the objective functions are crisp and the feasible region is a rough set and, in particular, discuss the properties of the completely and efficient (Pareto optimal) solutions of rough multiple objective programming problems. In order to obtain these solutions we need certain theorems which we derive. Finally, we illustrate our results by examples.

Effectiveness of root cause analysis efforts, following a control chart signal, will be enhanced if there exist more accurate information about the true time of the change in the process. In this study, we consider a Poisson process ex- periencing an unknown multiple number of step changes in the Poisson rate. We formulate the multiple change point scenario using Bayesian hierarchical models. We compute posterior distributions of the change point parameters including number, location and magnitude of changes and also corresponding probabilistic intervals and inferences through Reversible Jump Markov Chain Monte Carlo methods. The performance of the Bayesian estimator is investi- gated over several simulated change point scenarios. Results show that when the proposed Bayesian estimator is used in conjunction with the c-chart, it can provide precise estimates about the underlying change point scenario (number, timing, direction and size of step changes). In comparison with alternatives including Poisson EWMA and CUSUM built-in estimators and a maximum likelihood estimator, our estimator performs satisfactorily over consecutive monotonic and non-monotonic changes. The proposed Bayesian model and computation framework also benefit of probability quantification as well as flexibility which allows us to formulate other process types and change scenarios.

An important problem in todays industries is the cost issue, due to the high level of competition in the global market. This fact obliges organizations to focus on improvement of their production-distribution routes, in order to nd the best. The Supply Chain Network (SCN) is one of the, so-called, production-distribution models that has many layers and/or echelons. In this paper, a new SCN, which is more compatible with real world problems is presented, and then, two novel hybrid algorithms have been developed to solve the model. Each hybrid algorithm integrates the simulation technique with two metaheuristic algorithms, including the Genetic Algorithm (GA) and the Simulated Annealing Algorithm (SAA), namely, HSIM-META. The output of the simulation model is inserted as the initial population in tuned-parameter metaheuristic algorithms to nd near optimum solutions, which is in fact a new approach in the literature. To analyze the performance of the proposed algorithms, di erent numerical examples are presented. The computational results of the proposed HSIM-META, including hybrid simulation-GA (HSIM-GA) and hybrid simulation-SAA (HSIM-SAA), are compared to the GA and the SAA. Computational results show that the proposed HSIM-META has suitable accuracy and speed for use in real world applications.

Nowadays, Value –based Supply Chain Management (VbSCM) is considered as a resource of competitive advantages, and companies with a long term strategic plan find the VbSCM as an effective factor in sustainability. In this context, supply chain network design has a significant impact on all value drivers (i.e. sales, supply chain costs, fixed assets and working capital). This paper proposes a stochastic mixed integer linear programming model for the value-based supply chain network design in which decisions on physical flow (raw materials and finished products) and financial flow are integrated. The proposed model is designed for a four-echelon, multi-commodity, multi-period supply chain and maximizes the value of the company based on the economic value-added concept by making some strategic and tactical decisions affecting the value drivers. Furthermore, a scenario-based two-stage stochastic programming model is developed with a scenario generation method based on Nataf transformation. Also, a computational analysis is done to illustrate performance of the proposed approach.

A grey target decision-making model is an effective method used to look for a relatively optimal decision-making scheme. In this method, whether a scheme is good or bad is determined through comparing the square sums of the differences between the evaluated indices and the optimal indices. However, such “power operation” probably results in amplification or reduction of some extreme index values in decision-making results. In this paper, an improved method based on cobweb area is proposed. Here an index is represented by a line drawn from the bull’s-eye, with equal angles between adjacent lines. Then data points are determined on the lines so that the length of a line segment represents the size of the index value. Each point is then connected in order, and a cobweb-like geometrical figure is obtained. With the proposed figure, each scheme could be evaluated by finding the area of its corresponding cobweb. The proposed model was applied for choosing the preferred software development mode of Chana Group’s Office Automation system, and the new model’sperformance was then compared with that of the traditional grey target decision-making model. The comparison shows the new model is more outstanding than a traditional grey target model.

In this paper, we study scheduling of R&D projects in which activities may to be failed due to the technological risks. We consider two introduced problems in the literature referred to as R&D Project Scheduling Problem (RDPSP) and Alternative Technologies Project Scheduling Problem (ATPSP). In the both problems, the goal is the maximization of the expected net present value of activities where activities are precedence related and each of them is accompanied with a cost, a duration and a probability of technical success. In RDPSP, a project payoff is obtained if all activities are succeeded while in ATPSP, if one of activities is implemented successfully, the project payoff is attained. We developed a solution representation for each of these problems and developed two population-based metaheuristics including scatter search algorithm and genetic algorithm as solution approaches. Computational experiments indicate scatter search outperforms genetic algorithm and also available exact solution algorithms.

Determining the number of operator for manufacturing operations is important in Assembly line balancing.Optimal allocation of manpower increase production efficiency. Thus it increases profits of the companies. Since there is a possibility to assign different numbers of machines to each operator, therefore a variety of scenarios of machine assignment to operators will occur. Our goal in writing this paper is to help managers choose the best possible scenario. In this model, each of the possible scenarios are modeled using the principles of queuing theory and costs and revenues from each of these scenarios is calculated.Since uncertainty is an important part of the manufacturing environments, thus a fuzzy logic model is proposed to consider the uncertainty in problem. Since some input of the model such as service rate and arrival rate are fuzzy, thus profit of the model will be a fuzzy number. Therefore we use fuzzy ranking methods for prioritizing the scenarios.