在线多单位组合拍卖机制及在初始排污权分配中的应用研究

Based on the non-gratuitous allocation of initial emission permits currently carried out in China, this project studies the online multi-unit combinatorial auction mechanism, aiming at the network-based real-time auction of multiple units of different items.. Firstly, the description form of the asymmetric information, including preference types, arrival time and departure time of bidders, is given, and the specified bidding language is designed. The bidding language can provide an accurate quotation for multiple units of different items over time, which can help the auctioneer to separate all goods into several groups and reduce the risk of bidders.. Secondly, the constraints, including incentive compatibility, individual rationality, etc., are formulated. And then, the mechanism design model for online multi-unit combinatorial auctions is built, which can make real-time decisions. The objective of this model is the efficient allocation of resources. . Thirdly, the necessary and sufficient conditions for incentive compatibility and individual rationality are derived, the bidding pre-processing methods are designed, and the efficient allocation rule and the payment rule are proposed. Thus, the online multi-unit combinatorial auction mechanism is generated.. Finally, the online multi-unit combinatorial auction mechanism is applied into the allocation of initial emission permits, and further simulation experiments are conducted, so as to test the validity and effectiveness of the proposed model, algorithm and mechanism.. This project can enrich the mechanism design theory in online multi-unit combinatorial auctions, and help build the non-gratuitous allocation and trading system of emission permits.

以我国正在推进的初始排污权有偿分配为应用背景，针对基于网络的多数量单位、多种类物品实时拍卖的问题，研究在线多单位组合拍卖机制。. 为此，首先给出包含竞标人到达时间、离开时间及偏好类型的不对称信息描述形式，设计一种可随时对多单位异质物品进行报价的投标语言，促使拍卖人对物品合理划分，降低竞标人风险；其次，构造激励相容、个体理性等约束条件，以资源有效配置为目标，建立能实时决策的在线多单位组合拍卖机制设计模型；再次，推导出满足激励相容和个体理性的充要条件，设计投标预处理方法，获得有效分配规则及支付规则，构造出在线多单位组合拍卖机制。最后，将提出的在线多单位组合拍卖机制应用于初始排污权的分配问题中，通过仿真实验，验证模型的合理性、算法的高效性和机制的适用性。. 研究成果不仅能丰富和发展在线多单位组合拍卖机制设计理论，而且有助于排污权有偿使用和交易制度的建立。

随着网络技术和电子商务的飞速发展，在线拍卖以其方便快捷和经济性的优势引起了人们的广泛关注。诸如广告机会的销售、云资源的分配、初始排污权交易等在线拍卖实践中，待拍卖的往往是多种具有替代或互补关系的异质物品，且每种物品的数量大于1，同时买家对异质物品有特殊的组合需求，且一次拍卖中提出多种需求。因此，研究和探索上述市场特性的拍卖机制，可推动在线拍卖市场的健康有序发展。. 在理论层面上，本课题设计了一种可在任意时间对多种异质物品进行不同数量报价的投标语言，以卖家期望效用最大化为目标，建立了能实时决策的在线多单位组合拍卖机制设计模型，提出了具有单调性的分配函数和基于最低赢标估值信号的支付函数，从而得到了在线多单位组合拍卖机制。该研究结果丰富和发展了在线拍卖机制设计理论。. 在应用层面上，针对多数量单位、多种类排污权的初始分配问题，以社会福利最大化为目标，提出了允许排污企业提交多种需求的在线组合拍卖机制。通过引入排污权稀有度的参数，平衡了各排污权在分配过程中的消耗速度。实验结果表明，相比于完全信息下的有效分配，所设计机制在社会福利、企业成交率、排污权利用率等多个方面能取得良好的效果；当排污企业提出多种排污权组合需求时，所提出机制可增加社会福利。本课题所提出机制消除了时间和空间的限制，满足了排污企业对排污权的多样性需求，有助于公平有效地分配初始排污权。