---

Developing realistic models for the evaluation of sustainable supply chains has turned into a major challenge facing managers. The decision-making approaches proposed here consist of two stages. At the first stage, a dynamic-network data envelopment analysis (DNDEA) model is established for the first time, wherein the current efficiency of a business can be influenced by its prior social and environmental activities, as two main dimensions of sustainability. The second stage correspondingly presents, for the first time, a model in which total efficiency is calculated based on the value of historical data. Sensitivity analysis is exploited to determine the more effective factors of sustainability in efficiency evaluations. To validate the model, it is used to assess the sustainability of the suppliers of an auto spare parts manufacturer. The study results reveal that the model is well-able to evaluate the performance of dynamic network structures, with a very high discriminating power. Following the implementation of this model, only the supplier(KARAN) is found to reach the efficiency limit, and  SIRIN S.N. is recognized as the most inefficient supplier with an efficiency score of 0.6409. The sensitivity analysis outcomes demonstrate that the least amount of efficiency change is related to the economic pillar; however, the rising trend in wage costs, compared with other economic factors, brings a better effect on augmenting the efficiency of some inefficient suppliers. The highest efficiency changes during sensitivity analysis are further observed in both social and environmental dimensions. Therefore, it is claimed that investing in these two pillars can have a significant impact on the efficiency of suppliers.
 

---

Due to the importance of vehicle routing for delivering a large number of orders with different restrictions in the world, various optimization methods have been studied in past researches. In this article, a number of researches of recent years have been discussed, then the proposed model is described in 3 phases with the penalty index. This model has the ability to assign orders, route vehicles and determine the number of active vehicles dynamically with the aim of minimizing the total cost of distribution. By examining valid metaheuristic models and using their strengths and weaknesses, and considering multiple limitations, a new model of "dynamic 3-phase optimization" has been designed. The main application of the proposed model is for vehicle routing problems with capacity constraints of fleet number and capacity constraints (maximum and minimum number of orders). Finally, with simulation, the outputs of the model have been analyzed in different conditions . Although the limitation of maximum and minimum capacity is added to the problem, by dynamically considering the number of vehicles and using star clustering (initiative of this research), three social, environmental and economic dimensions were improved. The time for orders to reach customers decreased by 19.3%, fuel consumption and air pollution by 14.9%, and logistics costs by 8.7%. To calculate the final value of system stability, a unique 3D fuzzy model has been used. With the sensitivity analysis, we came to the conclusion that the 3-phase dynamic optimization model has led to a 14.58% improvement in system stability.