Research On The Winner Determination And Mechanism Design Problem Of EAR

With the increasing competition in the market, electronic reverse auction (ERA) is becoming a common choice of the government and the enterprise procurement. As a new procurement mode, ERA brings various opportunities to many companies, and to some extent, improves their competitiveness. However, without enough practical and theoretical experience, the diffusion of online reverse auction is limited and the result of using the ERA often defeats the expectation. ERA does have some advantages, but because of its novelty and complexity, many companies do not know or understand the ERA. Therefore, the research of main issues in the ERA is significant from the theoretical and practical aspects.This dissertation focuses on the supplier evaluation, the winner determination and the auction mechanism design problem for the ERA environment. Firstly, we do some literature review on what we care, and then summarize advantages, disadvantages and using constraints of the EAR. In this dissertation, we mainly present the results of supplier valuation, winner determination and multiattribute VCG mechanism design problem from the buyer’s aspect. The content of the dissertation is described as follows:(1) An overview of the ERA is presented in Chapter2. Firstly, the auction theory and its classification are described. And then, we introduce the significance and advantages of online auction, the purpose of procurement and the process of electronic procurement, and the using conditions of the ERA. It is essential to the following research.(2) The supplier evaluation problem is described in Chapter3. Firstly, we discuss the positive and negative factors impacting the results under the ERA circumstances. Then we use the BOCR-FAHP approach to evaluate the suppliers’biddings in the ERA. Finally, a numerical instance is proposed, and the result shows the effectiveness of the method.(3) The winner determination problem of single item multiple unit auction for the ERA environment is proposed in Chapter4. Firstly, we describe the general formulation of the problem, and then the supply mode is incorporated in the model. A heuristic algorithm is proposed to solve the two models, and finally, numerical instances are used to illustrate the effectiveness of the models and the methods.(4) The winner determination problem of multiple item single unit auction for the ERA environment is proposed in Chapter5. Firstly, a model considering the total procurement cost and the multiattribute value is formed and its simplified version is discussed. Then, the enumeration algorithm based on backtracking, the LEPF heuristic, the improved ant colony algorithm and the quantum particle swarm algorithm are proposed to solve the NP complete problem. Finally, numerical instances are used to show that the model and the methods are effective.(5) A simple auction mechanism of ERA is described in Chapter6. Firstly, we formulate the model to maximize the social welfare. Then, the payoff rule is proposed and the buyer-supplier repeated game model is described based on the former model. Finally, the mechanism is proved to be dominant-strategy incentive-compatible, that is, each bidder has incentive to bid truthfully. When the buyer’s preference is common knowledge, the mechanism is also budget-balanced.