在线商务的市场建模、分配与定价机制研究

发布时间：2018-03-15 03:00

本文选题：市场建模　切入点：博弈竞价　出处：《扬州大学》2017年硕士论文　论文类型：学位论文

【摘要】：目前,电子化市场已经成为极为重要的商品交换平台与方式之一。随着在线商务的飞速发展,市场模型及商品分配与定价机制逐渐成为学术界和产业界研究的热点内容。现有市场模型的研究主要集中在利益优化、价格构成和激励措施,缺少市场机制的设计、开发与效率评估等方面研究。本文在研究相关在线商务理论和模型的基础上,对在线商务市场进行建模,设计了带动态环境下自适应的妥协协商机制,通过多边多议题策略集最大逻辑一致求解与离散化算法对在线商务市场建模,并选用IJCAI的TAC竞赛提供的Market Design实验平台,验证本文提出方案对动态环境的处理能力以及策略调整对智能体的市场竞争力的影响。研究的主要内容及贡献包括:(1)设计了一个电子化市场模型,并给出市场形式化描述语言,进而阐述了带动态环境自适应的妥协协商机制及最大逻辑一致求解策略。(2)提出了一种在多智能体系统中,表征系统内部agent个体对赢者集的满意度计算方法。基于满意度,针对单一获胜者和多位获胜者两种情形,分别设计了 SWDP和CWDP算法。对照实验表明,SWDP和CWDP算法能够聚集多agent系统(MAS)中的个体意愿,并且选出具有最优满意度的赢者集。通过证明:SWDP算法满足帕累托公理。(3)提出一种伙伴利益均衡的求解方法,包括基于最大社会效益的分组算法和联盟内利益迭代分配算法。在介绍了形成伙伴或结盟、利益分配的相关工作之后,给出了问题模型和相关概念的形式化定义;分别设计了基于最大社会效益的分组算法和联盟内利益迭代分配算法,并给出了算法的性能分析,以市政供水规划为例详细阐述了伙伴利益均衡求解方法及过程。(4)研究了刚性价格市场下的拍卖机制,定义了潜在市场模型,并设计了动态拍卖程序,并证明了这个程序可以在有限的步骤内找到一个有效的竞争均衡,并给出了一个刚性价格限制下的拍卖机制运行的实例。以上的研究为电子商务的市场建模、分配及定价奠定了理论基础,在给出了电子化市场建模的基础上,分别研究了社会满意度最优,伙伴合作利益均衡和刚性价格条件下的竞争均衡等问题,通过算法及应用实例验证了本文提出机制的有效性。
[Abstract]:At present, the electronic market has become one of the most important commodity exchange platforms and methods. With the rapid development of online commerce, Market models and commodity distribution and pricing mechanisms have gradually become hot topics in academia and industry. The existing market models focus on interest optimization, price composition and incentive measures, and lack of market mechanism design. Based on the research of related online business theories and models, this paper models the online commerce market and designs an adaptive compromise negotiation mechanism with dynamic environment. The online business market is modeled by the maximum logical consistent solution and discretization algorithm of the multi-topic strategy set, and the Market Design experimental platform provided by IJCAI's TAC competition is selected. The main contents and contributions of this paper include the design of an electronic market model, and the formal description language of the market. Furthermore, the compromise negotiation mechanism with dynamic environment adaptive and the maximum logic consistent solution strategy. 2) A method to calculate the satisfaction degree of the agent individuals to the winner set in a multi-agent system is proposed, which is based on the degree of satisfaction. In this paper, SWDP and CWDP algorithms are designed for the case of single winner and multi-winner, respectively. The comparative experiments show that SWDP and CWDP can gather the individual will in multi-#en3# system. And the winner set with optimal satisfaction is selected. By proving that the SWDP algorithm satisfies Pareto axiom. It includes grouping algorithm based on maximum social benefit and iterative benefit allocation algorithm in alliance. After introducing the related work of forming partner or alliance and benefit distribution, the formal definition of problem model and related concepts is given. The grouping algorithm based on the maximum social benefit and the iterative benefit allocation algorithm in the alliance are designed, and the performance analysis of the algorithm is given. Taking the municipal water supply planning as an example, the paper expounds the method and process of solving the benefit balance of the partners in detail. It studies the auction mechanism in the rigid price market, defines the potential market model, and designs the dynamic auction program. It is proved that this program can find an effective competitive equilibrium in limited steps, and an example of auction mechanism running under rigid price constraints is given. Distribution and pricing have laid the theoretical foundation. On the basis of the electronic market model, we have studied the social satisfaction degree optimization, the partner cooperation benefit equilibrium and the competition equilibrium under the rigid price condition and so on. The effectiveness of the proposed mechanism is verified by an algorithm and an application example.
【学位授予单位】：扬州大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：F724.6;TP18

【参考文献】