Control and Optimization
Sajad Amirian; Maghsoud Amiri; Mohammad Taghi Taghavifard
Abstract
Integrating sustainability and reliability represents a synergistic approach that can be explored through the problem of a closed-loop supply chain network design (SCND). This study is conducted in three stages: mathematical modeling, model solution using exact methods, ...
Read More
Integrating sustainability and reliability represents a synergistic approach that can be explored through the problem of a closed-loop supply chain network design (SCND). This study is conducted in three stages: mathematical modeling, model solution using exact methods, and evaluation of the solution methods. In the first stage, a mixed-integer linear programming (MILP) model is developed in a multi-objective, multi-product, and multi-period framework. The objectives of the proposed model aim to maximize profitability, social responsibility, and reliability. In the second stage, two methods, namely Augmented $\varepsilon$-Constraint (AEC) and Normalized Normal Constraint (NNC), are implemented in the GAMS software to solve the model and identify the optimal Pareto solutions. In the third stage, the Shannon Entropy technique is employed to determine the criteria weights, and the VIKOR technique is utilized to select the superior solution method. The overall performance accuracy of the proposed model is measured using four samples from a numerical example with randomly generated data based on the objective function coefficients. The results indicate the presence of a conflict among the three objective functions. Consequently, decision-makers should consider sacrificing some profitability to enhance environmental protection and improve reliability. In terms of three criteria, run time, diversification metric, and general distance, the NNC method is given priority over the AEC method. Even when the criteria are given equal weight, the superiority of the NNC method remains unchanged. The application of the proposed model across different industries represents a significant research direction for future research.
Hamidreza Ayoughi; Hossein Dehghani Poudeh; Abbas Raad; Davood Talebi
Abstract
In this paper, a stable multi-objective model of location, inventory, and supply chain routing is presented under conditions of uncertainty and using a passive defense approach. Parameters such as demand, cost of setting up the facility ...
Read More
In this paper, a stable multi-objective model of location, inventory, and supply chain routing is presented under conditions of uncertainty and using a passive defense approach. Parameters such as demand, cost of setting up the facility and cost of maintaining inventory are considered uncertain and in the form of triangular fuzzy numbers. Also, in order to increase supply chain resilience, the characteristics and capabilities of passive defense in the supply chain, such as ``ready flow rate'', ``security of backup routes'', ``possibility of deployment of resources and equipment'', and ``the principle of dispersion for location'' are considered. Multipurpose, multipartite algorithms, based on the Pareto archive and genetic algorithm, are used to solve the model. The results of validation show that the proposed model is valid and feasible, and the proposed algorithm is also valid and converges to the optimal solution. Sample problems, in three groups of small, medium and large, are solved by two algorithms, and the results are compared based on quality, dispersion, uniformity and execution time. The results of this section show that in all cases, the multi-objective particle mass algorithm has a higher ability than the GA to produce solutions of higher quality and to explore and extract the scalable area of the solution. Also, the comparison of the execution times of the algorithms indicates that the multi-objective particle mass algorithm has a higher solution time.