استفاده از بهینه سازی ازدحام ذرات خطی برای مدل موجودی چند پله ای زنجیره ای

In the aspect of supply chain management; responding to the accurate needs of customers and effectively reducing the total costs are of significance activities for companies to achieve their competitive edges and to have opportunities to gain large amounts of advantages in the current highly competitive global supply chain management environment. Therefore, this article provides a serial multi-echelon integrated just-in-time (JIT) model based on uncertain delivery lead time and quality unreliability (SMEIJI model) considerations. Hence, we will apply the particle swarm optimization (PSO) as a method to result an improved solution solving a mixed nonlinear integer problem. Based on our fitting parameter settings, the final result show that the linear decreasing weight particle swarm optimization (LDW-PSO) will be efficiently performable and have a primary solution in solving the multi-echelon inventory problem.

The main discussion of this article is to create a serial multi-echelon integrated just-in-time inventory model with an uncertain delivery lead time and quality unreliability consideration; above all, we focus on comparing LDW-PSO with the non-liner decreasing weight particle swarm optimization (NLDW-PSO). Accordingly, Ha and Kim (1997) indicates that the just-in-time method attempts to eliminate all waste from a firm’s operation, and ultimately, help the firm achieve zero inventories by facilitating frequent shipment of purchased parts in small lots and manufacturing small lots frequently. Even though the performance of whole supply chain is above that of a single echelon member in this global competition environment, most companies still hardly to avoid this disadvantage. The possible contribution factor is that buyers may not adjoin vendors geographically in real situation. The uncertain delivery lead time might be happened and exposure to the safety inventory stock. We can say that, it is directly decreasing customer satisfaction and rising by higher defective rate. Hence, we have to find out an optimal inventory model that can ameliorate the total cost of whole supply chain under uncertain delivery lead time and quality unreliability. The key factor of an enterprise to gain advantage in supply chain management is to create an optimal inventory development. Especially, enterprises have trying to respond to the real requirements of all customers and reduce total costs effectively. In traditional inventory system, both vendors and buyers only focus their own optimal economic lot sizes; however, this does not obtain in an optimal policy for the entire supply chain. To be specific, an old-fashioned inventory policy might not be economically realistic in the current highly competitive global supply chain environment. Consequently, an integrated inventory approach may help determine an optimal order quantity and shipment policy. The structure of this article is indicated as follows: the material and method reviews of serial multi-echelon just-in time inventory, LDW-PSO and NLDW-PSO. And then, based on our proper parameter settings; the computational experiments will be performed with the particle swarm optimizations that have good initial solutions. Comparing the assumption results with the ideal solution found by LINGO 9.0 is our main discussion; and finally, we will display all computational results to have conclusions and suggestions.

To sum up, there are two main topics in optimal supply chain management; responding to the accurate requirements of customers and effectively reduce the joint total cost. Above all, the vendor–buyer relationships are becoming more inseparable recently; however, there are few studies over the last decade have attempted to solve the multi-echelon integrated inventory problem. As the result, this article expresses a serial multi-echelon integrated JIT inventory model with uncertain delivery lead time and quality unreliability, and considers the problem of finding an optimal policy in a serial multi-echelon integrated JIT inventory model. Since optimizing the proposed model is equivalent to solving a mixed nonlinear integer problem, this study uses PSO to propose a method of generating a good initial solution. Results show that PSO with a good initial solution is rapid and efficient in solving the multi-echelon inventory problem with proper parameter settings. And it is helpful of companies and researchers to extend their further studies.