A Research To The Closed-loop Supply Chains In Two E-commerce Modes

With the development of the E-Commerce, great changes have taken place in the market. Online shopping has been widely accepted by consumers which have promoted the development of multi-channel sales. In such a diversity shopping environment, the traditional supply chain network has gradually revealed its deficiencies. Consumers’ behavior has increasingly become the key factors to influence supply chain. Meanwhile, in the background of the sustainable development, reverse logistics to strengthen the recycling of products has become a major trend. Companies hope that through the application of reverse logistics to reduce their production costs and improve their economic benefits, but also in response to the call of Environmental protection and Energy saving. There are two models of E-Commerce in the current market. One evolves from traditional enterprise who sells products through both online and offline. The other is new enterprises that focus to on-line shopping only.This paper studies the traditional companies first. Now company’s sales channels are both online and offline. We formulate a closed-loop supply chain network help to make decisions of location selection and service distribution with consideration of consumers’ purchase manner. Moreover, we discuss the influence of the probability of different selections of consumers’ purchase manner on the supply chain network and give some price setting advices. In order to solve this problem with big scale, we propose a heuristic solution and test its effectiveness and efficiency with extensive numerical experiments.Then this paper studies the E- Commerce companies. The biggest difference of ECommerce companies with traditional companies is removing the production part. They obtain the finished product from suppliers and sell them directly. However, these companies also contain a large number of suppliers. We formulate a model to arrange the order point and retrieve point for the multiply suppliers and set a case study to test the practicability of the model.