Scientia Iranica
Volume:26 Issue: 6, Nov-Dec 2019

This paper presents a novel formulation of the integrated bi-objective problem of project selection and scheduling. The first objective is to minimize the aggregated risk by evaluating the expected value of schedule delay and the second objective is to maximize the achieved benefit. To evaluate the expected aggregated impacts of risks, an objective function based on the Bayesian Networks is proposed. In the extant mathematical models of the joint problem of project selection and scheduling, projects are selected and scheduled without considering the risk network of the projects indicating the individual and interaction effects of risks impressing the duration of the activities. To solve the model, two solution approaches have been developed, one exact and one metaheuristic approach. Goal Programming method is used to optimally select and schedule projects. Since the problem is NP hard, an algorithm, named GPGA, which combines Goal Programming method and Genetic Algorithm is proposed. Finally, the efficiency of the proposed algorithm is assessed not only based on small size instances but also by generating and testing representative datasets of larger instances. The results of the computational experiments indicate that it has acceptable performance to handle large size and more realistic problems.

Nowadays, manufacturing environments are faced with globalization which urges new necessities for manufacturing systems. These necessities have been considered from different perspectives and Computer Integrated Manufacturing (CIM) is the most popular and effective one. However, considering rapid rate of manufacturing globalization, traditional and current CIM solutions can be criticized by major deficiencies like high complexity for resource allocation over the globe, global facility sharing, and absence of an efficient way to handle lifecycle issues. Recently, Virtual CIM (VCIM) has been introduced as an effective solution to extend the traditional CIM solutions. This paper has investigated recent researches in VCIM/CIM field considering the necessities of todays’ globalized manufacturing environment. The paper shows the lack of traditional and current CIM/VCIM solutions; then, proposes an effective solution to cover them. Because of the complexities in designing such systems, the paper exploits Axiomatic Design (AD) Theory as a promising tools in this field. This theory is applied for validation of the suggested architectural solution and verification of the implementational aspects. The implementation of the architectural solution is considered based on ISO standards. Finally, the results have approved the feasibility of the suggested solution for manufacturing system and its Implementation aspects.

Banks need to open new branches in new sites as a result of increase in the population, individual earnings and the growth in national economy. In this respect, opening new branches or reorganizing the locations of current branches is an important decision problem for banks to accomplish their strategic objectives. This paper presents a decision support method for multi-period bank branch location problems. Our aim is to find bank branch location based on transaction volume, distance between branches, and cost of opening and closing branches. The proposed method not only develops an Integer Program and a Tabu Search algorithm to find the exact places of branches but also presents a structuring method to identify the related criteria and their importance. We demonstrate the effectiveness of the method on random data. In the final stage, the method is applied in a Turkish bank’s branch location problem considering the current and possible places of the branches, availability of the data, and the bank’s strategies for a four-year strategic planning.

This study aims in providing a new approach regarding design of a closed loop supply chain network through emphasizing on the impact of the environmental government policies based on a bi-level mixed integer linear programming model. Government is considered as a leader in the first level and tends to set a collection rate policy which leads to collect more used products in order to ensure a minimum distribution ratio to satisfy a minimum demands. In the second level, private sector is considered as a follower and tries to maximize its profit by designing its own closed loop supply chain network according to the government used products collection policy. A heuristic algorithm and an adaptive genetic algorithm based on enumeration method are proposed and their performances are evaluated through computational experiences. The comparison among numerical examples reveals that there is an obvious conflict between the government and CLSC goals. Moreover, it shows that this conflict should be considered and elaborated in uncertain environment by applying Min-Max regret scenario based robust optimization approach. The results show the necessity of using robust bi-level programming in closed loop supply chain network design under the governmental legislative decisions as a leader-follower configuration.

Train formation planning faces two types of challenges; namely, the determination of the quantity of cargo trains run known as the frequency of cargo trains and the formation of desired allocations of demands to a freight train. To investigate the issues of train makeup and train routing simultaneously, this multi-objective model optimizes the total profit, satisfaction level of customers, yard activities in terms of the total size of a shunting operation, and underutilized train capacity. It also considers the guarantee for the yard-demand balance of flow, maximum and minimum limitations for the length of trains, maximum yard limitation for train formation, maximum yard limitation for operations related to shunting, maximum limitation for the train capacity, and upper limit of the capacity of each arc in passing trains. In this paper, a goal programming approach and an L p norm method are applied to the problem. Furthermore, a simulated annealing (SA) algorithm is designed. Some test problems are also carried out via simulation and solved using the SA algorithm. Furthermore, a sample investigation is carried out in a railway company in Iran. The findings show the capability and performance of the proposed approach to solve the problems in a real rail network.

Many researchers and practitioners in the recent years have been attracted to investigate the role of uncertainties in the supply chain management concept. In this paper a multi-period stochastic supply chain with demand uncertainty and supplier disruption is modeled. In the model, two types of retailers including risk sensitive and risk neutral, with many capacitated suppliers are considered. Autonomous retailers have three choices to satisfy demands: ordering from primary suppliers, reserved suppliers and spot market. The goal is to find the best behavior of the risk sensitive retailer, regarding the forward and option contracts, during several contract periods based on the profit function. Hence, an agent-based simulation approach has been developed to simulate the supply chain and transactions between retailers and unreliable suppliers. In addition, a Q-learning approach (as a method of reinforcement learning) has been developed to optimize the simulation procedure. Furthermore, different configurations for simulation procedure are analyzed. The R-netlogo package is used to implement the algorithm. Also a numerical example has been solved using the proposed simulation-optimization approach. Several sensitivity analyzes are conducted regarding different parameters of the model. Comparison of the numerical results with a genetic algorithm shows a significant efficiency of the proposed Q-leaning approach.

In this paper, we propose new Shewhart-EWMA (SEWMA) and Shewhart-CUSUM (SCUSUM) control charts using the varied L ranked set sampling (VLRSS) for monitoring the process mean, namely the SEWMA-VLRSS and SCUSUM-VLRSS charts. The run length characteristics of the proposed charts are computed using extensive Monte Carlo simulations. The proposed charts are compared with their existing counterparts in terms of the average and standard deviation of run lengths. It is found that, with perfect and imperfect rankings, the SEWMA-VLRSS and SCUSUM-VLRSS charts are more sensitive than their analogous charts based on simple random sampling, ranked set sampling (RSS) and median RSS schemes. A real dataset is also used to explain the implementation of the proposed control charts.

We present an integrated data envelopment analysis (DEA) and Malmquist productivity index (MPI) to evaluate the performance of decision making units (DMUs) by using a directional distance function with undesirable interval outputs. The MPI calculation is performed to compare the efficiency of the DMUs in distinct time periods. The uncertainty inherent in real-world problems is considered by using the best and worst-case scenarios, defining an interval for the MPI and reflecting the DMUs’ advancement or regress. The optimal solution of the robust model lies in the efficiency interval, i.e., it is always equal to or less than the optimal solution in the optimistic case and equal to or greater than the optimal solution in the pessimistic case. We also present a case study in the banking industry to demonstrate applicability and efficacy of the proposed integrated approach.

In survey sampling, most of the research work based on the fact that utilizing the information of auxiliary variable(s) boosts the efficiency of estimators. Keeping this fact in mind we used the information of two auxiliary variables to propose a family of Hartley-Ross type unbiased estimators for estimating population mean under simple random sampling without replacement. Minimum variance of the new family is derived up to first order of approximation. Three real data sets are used to verify that the new family acts efficiently than the usual unbiased, Hartley and Ross (1954), Grover and Kaur (2014), Singh et al. (2014), Cekim and Kadilar (2016), Muneer et al. (2017) and Shabbir and Gupta (2017) estimators.

Under restructuring of electric power industry and changing traditional vertically integrated electric utility structure to competitive, market clearing price (MCP) prediction models are essential for all generation company (GenCos) for their survival under new deregulated environment. In this paper, a hybrid model is presented to predict hourly electricity MCP. The model contains a Neural Network (NN), Particle swarm optimization (PSO) and Genetic Algorithm (GA). The NN is used as the major forecasting module to predict the electricity MCP values and PSO applied to improve the traditional neural network learning capability and optimizing the weights of the NN and GA applied to optimize NN architecture. The main contribution includes: presenting a hybrid intelligent model for MCP prediction; applying K-Means algorithm to clustering NN’s test set and seasonality pattern detection; and evaluation of its performance by improved MAE with penalty factor for positive error. It has been tested on Iranian real-world electricity market for the one month of the year 2010-2013 that result shown average weighted MAE for day ahead MCP prediction is equal to 0.12 and forecasting of MCP can be improved by more than 6.7% and 4%in MAE in compare of simple NN and combination of NN and bat algorithm.