Abstract: Because the existing multi-unit reverse multi-attribute auction mechanisms are low efficiency of the market allocation, they are not conducive to maximization of social welfare and long-term cooperation between buyers and sellers. To resolve the above problems, based on the strategy of price discrimination we design a highly efficient Multi-unit Reverse Multi-attribute English Auction Mechanism(MRMEAM), build the Winner Determination model using linear mixed-integer programming and give the auction process and bidding strategies for suppliers bidding decision support. Then we collect data through the design of numerical example and simulation comparing experiment, and we verify the new auction mechanism’s market allocation efficiency and utility of the purchaser are both higher by statistical analysis method. The above study results are applicable to online automatic multi-unit procurement through the combination of reverse multi-attribute auction and information technology.

Key words: multi-unit reverse multi-attribute auction, auction mechanism, linear mixed-integer programming, price discrimination

摘要： 针对现有的多单元逆向多属性拍卖机制的市场分配效率都比较低,不利于社会效益最大化和采购双方长期合作等问题,设计了高效率的基于价格歧视策略的多单元逆向多属性英式拍卖机制,利用线性混合整数规划建立了赢者确定模型,并基于新建立的拍卖机制给出了拍卖流程和投标策略,为供应商提供投标决策支持。然后通过设计数值实例和对比模拟实验收集数据,采取统计分析的方法验证了新设计的拍卖机制市场分配效率和采购方的效益都比较高。上述研究结果适用于通过逆向多属性拍卖与信息技术相结合的网上自动化多单元商品采购。

关键词: 多单元逆向多属性拍卖, 拍卖机制, 线性混合整数规划, 价格歧视