The Mechanism Design Of Bi-attribute Electronic Reverse Auction

With the development of network economy and information technology, e-commerce has been developed rapidly. Nowadays, e-commerce has become an indispensable part in our daily life and the national economy as a whole. As an important application field of e-commerce, electronic reverse auction (ERA) has been widely used. It breaks the limitation of time and area about traditional purchasing mode, enlarges the range of the suppliers who can participate in the procurement, and improves competitiveness of enterprises to some extent. However, many enterprises and governments often cannot gain the expected results when apply it, due to its novelty and complexity and a lack of sufficient practical and theoretical experience. In addition, because of the diversification of the market as well as the multiplicity of goods and services, price-only ERA cannot satisfy the actual requirement from the buyer. Therefore, it is important and valuable to do the research about mechanism design of bi-attribute ERA with the consideration of price and delivery time both in theory and in practice.Based on auction theory, reverse auction theory and mechanism design theory, this paper makes a detail research on the ERA with price and delivery time from three aspects:the lowest quotation of suppliers is known, the lowest quotation of suppliers is unknown where the delivery time is decided by the purchaser, and the lowest quotation of suppliers is unknown where the delivery time is decided by the suppliers. The content of the research is mainly presented as follows:(1)Firstly, this paper reviews auction theory, and summarizes the differences between traditional auction and online auction. Secondly, ERA is introduced, including its process, merit and demerit and conditions of successful implementation. Finally, we summarize the basic theory of mechanism design, the research framework of centralized decision-making and distributed decision-making, and the basic ideas of negotiation. It is essential for the following research.(2) When the lowest quotation of suppliers is known, we make a detailed research on the bi-attribute ERA with price and delivery time. Based on the two characteristics of the lowest quotation with quantity discount and time discount, the centralized decision-making model is established. An enumerative algorithm is proposed to solve the model, and then a numerical example is applied to illustrate the effectiveness of the model and the algorithm. Here, the result under the circumstance of symmetric information in this part is an evaluation criterion under the circumstance of asymmetric information in the following two parts.(3) When the lowest quotation of suppliers is unknown and the delivery time is decided by the purchaser, the mechanism design of bi-attribute ERA with price and delivery time is studied. According to the problem where demanded quantity and delivery time is decided by purchaser and quotation is decided by suppliers, the bi-level decision model based on negotiation is established. The up-level model indicates the decision-making process of purchaser, while the down-level model is the decision-making process of suppliers. Due to the privacy of the delivery time, this part of the paper proposes three different strategies:they are the guiding strategy based on delivery date deviation, the guiding strategy based on objective value and the guiding strategy based on objective value with random. After that, a numerical example is applied to show the effectiveness of these strategies. Finally, based on the 50 randomly generated examples, three strategies are contrasted and analyzed, and the result shows that the third strategy is the best.(4) When the lowest quotation of suppliers is unknown and the delivery time is decided by the suppliers, the mechanism design of bi-attribute ERA with price and delivery time is studied. Considering the condition of quantity decided by purchaser as well as quotation and delivery time decided by suppliers, the bi-level decision-making model based on negotiation is proposed. Then three strategies are proposed:that are the guiding strategy based on delivery time (Strategy 1), the guiding strategy based on objective value (Strategy 2) and the guiding strategy based on objective value with random (Strategy 3). After that, a numerical example is applied to show the effectiveness of these strategies. At last, based on the 50 randomly generated examples, three strategies are compared with each other, and we conclude that Strategy 3 is better than Strategy 2, and thus is better than Strategy 1.