Mohammad Mohammadi; Kamran Forghani
Abstract
The cell formation problem and the group layout problem, both are two important problems in designing a cellular manufacturing system. The cell formation problem is consist of grouping parts into part families and machines into production cells. In addition, the group layout problem is to find the arrangement ...
Read More
The cell formation problem and the group layout problem, both are two important problems in designing a cellular manufacturing system. The cell formation problem is consist of grouping parts into part families and machines into production cells. In addition, the group layout problem is to find the arrangement of machines within the cells as well as the layout of cells.In this paper, an integrated approach is presented to solve the cell formation, group layout and routing problems. By Considering the dimension of machines, the width of the aisles, and the maximum permissible length of the plant site, a new framework, called spiral layout, is suggested for the layout of cellular manufacturing systems. To extend the applicability of the problem, parameters such as part demands, operation sequences, processing times and machine capacities are considered in the problem formulation. The problem is formulated as a bi-objective integer programming model, in which the first objective is to minimize the total material handling cost and the second one is to maximize the total similarity between machines. As the problem is NP-hard, three metaheuristic algorithms, based on Genetic Algorithm and Simulated Annealing are proposed to solve it. To enhance the performance of the algorithms, a Dynamic Programming algorithm is embedded within them. The performance of the algorithms is evaluated by solving numerical examples from the related literature. Finally, a comparison is carried out between the proposed spiral layout and the linear multi-row layout which has recently presented in the literature
Hossein Shams Shemirani1; Mahdi Bashiri; Mohammad Modarres
Abstract
In this research, optimization of examinations' timetable for university courses, based on a real problem in one of the universities in Iran is studied. The objective function defined for this problem is more practical and realistic than the other objective functions that have been utilized by previous ...
Read More
In this research, optimization of examinations' timetable for university courses, based on a real problem in one of the universities in Iran is studied. The objective function defined for this problem is more practical and realistic than the other objective functions that have been utilized by previous researchers in literature and effectively reflects the real objective of the problem. In order to define the objective function, we have made use of Coulomb's law in electricity that says the magnitude of the electrostatic force of interaction between two point charges is directly proportional to the scalar multiplication of the magnitudes of the charges and inversely proportional to the square of the distance between them. We have defined a repulsive force between any pair of Examinations. The optimum solution is achieved when the sum of all forces is minimized. Hence, the obtained mathematical model is a non-linear programming with binary variables, similar to the quadratic assignment problem (QAP) which is an NP-Hard problem. This sort of problems can be solved exactly only if they are in small sizes. For solving this problem in medium and large scale, some methods are used based on Simulated Annealing (SA) algorithm and Imperialist Competitive algorithm (ICA). These algorithms can reach good sub-optimal solutions in a short period of time. Practical results of this mathematical model are already used in one of the national universities in Iran. The practical results demonstrate the high efficiency and effectiveness of this model.
Bahman Naderi
Abstract
In this paper hybrid flowshop scheduling problem where some jobs, not all, have to follow no-wait restriction (that is, the operations of that job must be processed with no stop) is examined. In the literature, all papers assume that all jobs of the shops have to follow no-wait restrictions. First, this ...
Read More
In this paper hybrid flowshop scheduling problem where some jobs, not all, have to follow no-wait restriction (that is, the operations of that job must be processed with no stop) is examined. In the literature, all papers assume that all jobs of the shops have to follow no-wait restrictions. First, this paper mathematically formulates the problem with two different mixed integer linear models under proposed considerations. The models are evaluated using two performance measures of size complexity and computational complexity. The small instances of the problem are solved using commercial software of mathematical programming. To solve larger instances of problem, two solution algorithms have been developed. These two algorithms are based on imperialist competitive algorithm and simulated annealing. A comprehensive numerical experiment including small and large instances is conducted to evaluate the models and algorithms. The results show that the imperialist competitive algorithm outperforms simulated annealing
Mehdi Yazdani; Mostafa Zandieh; Reza Tavakkoli-Moghaddam
Volume 12, Issue 33 , July 2015, , Pages 43-74
Abstract
In this paper, the dual-resource constrained flexible job-shop scheduling problem (DRCFJSP) with objective of minimizing the makespan is investigated. Under studied problem is NP-hard and mainly includes three sub-problems. The first one is to assign each operation to a machine out ...
Read More
In this paper, the dual-resource constrained flexible job-shop scheduling problem (DRCFJSP) with objective of minimizing the makespan is investigated. Under studied problem is NP-hard and mainly includes three sub-problems. The first one is to assign each operation to a machine out of a set of capable machines, the second one is to determine a worker among a set of skilled workers for processing each operation on the selected machine and the third one deals with sequencing the assigned operations on the machines considering workers in order to optimize the performance measure. In this paper, we provide a mathematical model for this problem and then propose a hybrid meta-heuristic algorithm for solving the problem. The proposed hybrid algorithm uses variable neighborhood search and simulated annealing algorithms to search in the solution space. Computational study with randomly generated test problems is performed to evaluate the performance of the proposed algorithm. The results show the proposed algorithms are effective approaches for solving the DRCFJSP.
Esmaeil Mehdizadeh; Vahid Rahimi
Volume 13, Issue 37 , July 2015, , Pages 123-159
Abstract
Abstract
This paper presents a mathematical model for solving dynamic cell formation problem, operator assignment and the inter-cellular and intra-cellular layouts simultaneously. The proposed model includes three objectives, the first objective seeks to minimize inter and intra-cell part movement, ...
Read More
Abstract
This paper presents a mathematical model for solving dynamic cell formation problem, operator assignment and the inter-cellular and intra-cellular layouts simultaneously. The proposed model includes three objectives, the first objective seeks to minimize inter and intra-cell part movement, machine relocation, second objective minimize operator related cost, third objective maximize ratio of consecutive forward flows. The model is Multi-objective; therefore, the LP-metric approach is used to solve it. In order to validate the model, the proposed model has been solved by using Lingo software. Then, due to NP-hardness of the cell formation problem, for solving large scale problems, a multi-objective simulated annealing algorithm proposed. Several numerical examples solved by Lingo software and multi-objective simulated annealing algorithm. Results show that the proposed multi-objective simulated annealing algorithm solved considerably time less than the software Lingo and also none of the answers obtained by the two methods are not dominated
