The Design Of Product Returns Reverse Logistics Network Under Stochastic Environment

With the development of the global economic integration, the competition among different enterprises' supply chain has become the main aspects of the competition. Like the normal positive logistics, the reverse logistics is also a supporter of supply chain system. But because it's harder to control the reverse flow of goods, the cost accounting, inventory management and network design of reverse logistics etc are relatively more complex than positive logistics, and the supply chain system management is more difficult with the joint of reverse logistics. While the profits of traditional positive logistics are being exploited to be saturated, the requirements of resources and environmental protection, the enhancement of competition between enterprises and the protection of consumers' rights are proclaiming the importance of reverse logistics. Especially for the manufacturing enterprises are concerned, products return is an inevitable process of the modern commercial activities, as well as the market competition is spurring the enterprises to continuously relax the products return limits so as to ensure the customer satisfaction and their product market share. Therefore types and quantities of returned and exchanged products surge, and if the reverse logistics operation and network design and management is greatly influence the economic benefits and sustainable development of these manufacturing enterprises to a large extent. In addition, manufacturing enterprise can optimize its resources configuration, achieve the goal of environment protection and resource conservation at the same time of making benefits when they design the supply chain with consideration of both traditional positive logistics and reverse logistics.There are two main aspects in logistics network design problem:network nodes localization and configuration problems and the path optimization selection among all these nodes. The commercial product returns reverse logistics network studied in this paper contains three kinds of nodes:customer nodes, initial recovery points for retruns and centralized recovery processing centers. Design the flow of returns reverse logistics as:firstly consumers return those returns to one of the original recovery points, then the returned goods are transported to a centralized district recovery center after reaching a certain number, in the recovery center the goods are tested and sorted intensively, and finally most returns return back to original sales market.Since the amount of customer returns cannot be known precisely in advance, the paper researches the product returns reverse logistics network under stochastic environment, and establishes a mathematical model of single product returns handling and management network for manufacturing enterprises from the angle of third--party logistics services providers. The modeling uses a combined method of the Expected Value Model (EVM) and Chance Constrained Model (CCP), Lingo optimization software tests that the mathematical model is a mixed integer programming model and its validity is also tested.Because of the NP-hard attribute of this problem, the genetic algorithms (GA) is developed, and the C++is designed to realize the algorithm for solving a problem with simulated data. The result is compared with the Lingo solution, and we find that this algorithm proposed by this paper has good calculation stability and reliability in solving the kind of above-mentioned problem.