Optimization and Disruption Mitigation in a Supply Chain

Today supply chains play an important role in determining the position of companies in their competitive environment. In this research, a three-level multi-products supply chain was modeled to minimize the total cost. Then, the mathematical model analyzed by considering disruptions based on three deferred sales policies: back orders, lost sales and outsourcing. Model solved efficiently by MATLAB software. The results show that the proposed algorithm of this research can neutralize the effect of the disruptions and cause a significant reduction in total cost of the system. The model is useful for helping decision makers to adopt an active approach to maintaining business benefits when disruptions take place in the supply chain.

The survival of organizations is due to their ability to deliver better outlets at a wider range, lower cost than competitors, faster performance, and provide the desired outputs (from a variety of aspects, such as quality, performance, cost, innovation, and delivery). Delivering those outputs depends on the ability of the organization to manage the flow of materials, information and money inside and outside the organization. This flow is known as the supply chain. In today's competitive market, the supply chain coordination (SCC) is of great significance. In the lack of coordination, the SC members optimize their own decisions independently. Supply chains may be complex and long and involve a large number of different business partners and their problems and issues are inevitable. Failure to resolve these problems will result in customers' dissatisfaction and loss of sale and impose lots of costs to organizations. Companies in the world-class owe many of their successes to supply chain management, hence the supply chain has become a popular management discourse in recent decades. Since supply chain loops are independent entities of economics, creating a mechanism that can coordinate the activities and objectives of these entities in order to optimize the overall system performance is a key issue in the management of the supply chain. Generally, traditional researches in supply chain focused on making decisions in the ideal conditions, where demand was defined, and the producer had complete information about the market; but in the real world, the complete market information is rarely available and obtaining Demand distributions may be hard and even impossible. In a dynamic and complex environment of the supply chain, uncertainty is considered as an important concern. The uncertainty of demand is also an important issue in the world of science and industry. These uncertainties of demand posed major challenges to supply chain management, leading to a reduction in real income. In order to take uncertainty and its effects into account, it is necessary to pay attention to the sources of uncertainty, which are disruptions or interruptions in all parts of the supply chain.

Many studies have been carried out on the main field of this research, the supply chain; many of the existing studies carried out in this field are qualitative articles, and there aren’t any quantitative modeling issues. Although such studies play a major role in explaining new topics or in learning the key concepts, their application is restricted to the above, and is less applicable to solving problems that impress competition of supply chains. Qualitative studies with quantitative statistical approaches are another category of research in the supply chain, although these researches have pretended quantitative due to the use of statistical tools, but the principles of research are based on qualitative concepts such as chain performance measurement, supply chain culture, supply chain interoperability, supply chain stability, lean supply chain, agile supply chain, and so on. Quantitative studies in the supply chain are the third category of research in the supply chain, which the present study is in this category, and attempts to simulate real supply chains and solve problems and concerns of different parts of chains. Due to the extensive research, this section can be divided into three main sub-domains: a) studies focused on demand forecasting and inventory control; b) pricing-focused studies; and c) studies focused on disruptions and uncertainties. Each of these sub-domains also deals with the study of two-level, three-level, and etc. To clarify the perspective of each subcategory, we will review the relevant literature.

1. Studies focused on demand forecasting and inventory control: In recent research, Rezayi's research has been based on the principles of artificial intelligence science and data mining to provide models for forecasting sales of the next month. The proposed models of this study are of a hybrid type and include steps 1- The reduction of dimension 2- clustering and 3- prediction. Although the results indicate the higher accuracy of these models compared to traditional models of prediction, there are limitations such as limited time intervals. In inventory control, one of the most important initial tasks in the field of multi-level inventory systems, with emphasis on the concept of "level inventory", which deals with the analysis of the cost function of multilevel systems, is Clarke and Scarf's paper, "Optimal Policies for Multi-Level Inventory Issues". In another study, it has been forecasted using the Bayesian network of gasoline in Iran. The model presented in this study has been expanded with the use of annual data for the years 1347-1386 and GDP, vehicle numbers, population and actual price of gasoline. The results of this research show that the Bayesian network has less error than the regression method.

2. Pricing-focused studies, uses concepts such as game theory to find the best pricing strategy.

3. In studies focused on distortions and uncertainties, the goal is to design supply chain network and develop models that are both acceptable and efficient in both normal and disruptive situations.

A review of the literature indicates that most of the available articles ignore the potential for breakdowns and disruptions in the supply chain, and generally assume the demand and supply parameters in a definitive manner. Other research in the literature has often overlooked the severity of the disruption and often focuses solely on complete disruption and few studies have focused on minor disruption; A minor disorder is also modeled definitively in.