相关论文

基于集成模拟的[CQVIP免费论文网]电子商务协同工作机制研究,电子商务[CQVIP免费论文网]谈文范文,电子商务[CQVIP免费论文网]谈文如何写,电子商务[CQVIP免费论文网]谈文如何写大纲

时间: 2016-8-20 7:17:45    作者: 佚名    来源:中国论文    查看:

企业通过协同工作,能整合资源,降低成本,提高产出。本文集成多智能体模拟和其他建模方法为技术手段,研究电子商务服务链中竞争与协同决策工作机制、电子商务服务商联盟协同管理工作机制以及电子商务服务商内角色协同事务工作机制。阐述了面向电子商务协同工作机制的集成模拟的必要性和实现原理。通过分析电子商务服务商间...

企业通过协同工作,能整合资源,降低成本,提高产出。本文集成多智能体模拟和其他建模方法为技术手段,研究电子商务服务链中竞争与协同决策工作机制、电子商务服务商联盟协同管理工作机制以及电子商务服务商内角色协同事务工作机制。阐述了面向电子商务协同工作机制的集成模拟的必要性和实现原理。通过分析电子商务服务商间和服务商内的协同工作过程,阐述使用集成模拟的必要性。在总结一般集成模拟模式的基础上,描述了电子商务协同工作机制集成模拟思路。通过对比分析多智能体模拟模型实现平台,给出基于Repast J 3.0,用Java编程实现本文所用集成模拟模型的思路,也讨论了基于系统调试的模型验证方案,以及定量和定性相结合的确认方法。集成多智能体和优化算法,对服务链中竞争与协同决策工作机制进行研究。构建了由门户接入服务商、产品服务提供商驱动模式下的竞争机制和协同决策优化模型。为求解这类多阶段、多周期的优化模型,基于传统遗传算法,设计一种具有较快寻优能力的Agent进化算法(AEA)。在这些工作基础上,设计集成多智能体和优化算法的系统框架,并编程实现该系统。针对模拟数据,基于博弈视角分析了不同定价策略和不同风险偏好下各企业最优决策。
结果表明:一般情况下,协同机制优于竞争机制,能产生高利润。而通过设计合理的约束机制,能保证协同行为稳定存在也能有效降低获利风险。集成元胞自动机和进化博弈对服务商联盟协同管理工作机制进行研究。设计成本和惩罚固定的电子商务服务商联盟群体博弈得益矩阵,对群体进化稳定策略进行复制动态分析。考虑工作个体的个性决策特征,设计基于历史信息和个体决策特性的混合学习算法,将其集成进入元胞自动机上的多智能体中,并编程实现模拟系统。实验结果表明:联盟规模对宏观工作趋势影响小,具有不同决策特性比例个体的联盟搭配、不同协同率、不同沟通方式和惩罚参数对群体行为的影响各异。集成复杂网络和进化博弈对服务商联盟协同管理工作机制进行研究。考虑成本分摊的群体博弈情形,对群体进化稳定策略(ESS)进行复制动态分析。考虑成本和惩罚都分摊,且工作个体决策特征各异,设计基于历史信息和个体决策特性的混合学习规则,并将其集成进入复杂网络上的多智能体中,并编程实现模拟系统。模拟实验结果表明:联盟规模对宏观工作趋势影响小,但对合作频率影响大;不同的得益矩阵参数设置、不同的工作难度和具有不同决策特性比例个体的联盟搭配对群体行为的影响各异。集成多智能体和离散模拟方法,探讨了协同工作环境下,动态任务与员工协同事务工作行为的互动影响问题。构建基于Agent的角色-任务匹配的多目标评价模型,设计面向离散动态任务达到的最小匹配和贪婪匹配角色-任务匹配算法,并编程实现模拟系统。通过比较,可知最小匹配法优于贪婪匹配法。基于最小匹配算法的实验结果表明:管理者偏好对任务执行的影响小于对个体能力增长的影响;不同专才通才搭配下的团队对任务的执行影响不同;协同学习率与员工能力增长正相关;不同的任务动态性对员工能力增长度的影响各异。
【关键词】 电子商务; 协同工作; 集成模拟; 多智能体模拟; Agent进化算法; 进化博弈; 离散模拟;
 Enterprises can integrate resources, reduce costs and improve performance through collaborative work. This paper employ integrated agent-based simulation and other simulation methods to explore the competitive and the collaborative work mechanism of e-commerce service chains, the collaborative management mechanism of e-commerce service provider coalitions, and collaborative affair work mechanism inside an e-commerce enterprise.The necessity and implementation how-toes of integrated simulation for collaborative work mechanism among e-commerce enterprises are described. By analyzing collaborative work process of intra-enterprise and inter-enterprise, we explain necessity of integrated simulation for special application. After concluding the general simulation methods, we present integrated simulation principles for collaborative work mechanisms of e-commerce enterprises. And then, implementation method of integrated simulation models in this paper are put forward and compared with common implementation platforms concerning agent-based simulation. The implementation uses Java to program the simulation models based on Repast. We also discuss how to verify the model by debugging and how to carry out qualitative and quantitative validation.The integrated simulation for e-commerce service chain’s competitive and collaborative mechanisms is proposed based on multi-agents and optimal algorithm. The main idea of the proposed approach is based on a multi-agent system with optimal profit of the pull, push, and collaborative models among the portal access service provider (PASP), the product service provider (PSP), and the mobile service provider (MSP). To resolve all these modes, the agent evolution algorithm (AEA) is proposed based on traditional genetic algorithm. Furthermore, on basis of these, we design an integrated system framework with multi-agents and AEA. To examine the feasibility of the framework, a prototype system is implemented. As for simulation data, we try to mine there for an appropriate decision for each entity with regards to variety pricing strategies and risk preferences from game theoretic perspective. It is found that in the situation where a collaborative mechanism is applied rationally, the profit of players is higher as compared to the other two situations where a competitive mechanism is employed. If some constraints are applied, the collaborative state will exist and be stable, and the risk will be kept at a low level.The integrated simulation for collaborative management mechanism of e-commerce service provider coalition is studied based on cellular automata and evolutionary game. By assigning payoff matrix with fixed cost and punishment, we apply the dynamic equation to analyze for an evolutionary stable strategy for the group in coalition. By setting evolution learning rules including historical information and neighbor characters and considering personal decisional characteristic, the simulation system is implemented. Simulation results show that the different coalition scale have minor effect on macro trend of group behavior, and different individual distributions as well as different communication modes and punishment parameters have varying effect on group behavior.The integrated simulation for collaborative management mechanism of e-commerce service provider coalition is researched based on complex network and evolutionary game. By designing payoff matrix with payoff-shared and punishment-shared, we model replicate the dynamic equation to analysis evolutionary stable strategy of the group in coalition. By applying evolution learning rules including historical information and neighbor decision characters, the simulation system is implemented. Simulation results indicate that the scale of coalition have minor effect on cooperation trend and major effect on cooperation frequency of group, the different parameters of payoff matrix, the different group that are made up of individuals with variety decision-making characteristic, and the work difficulty have varying effect on group behavior.The integrated simulation with agent-based simulation and discrete simulation is applied to explore how collaborative work behavior interacts with dynamic tasks under collaborative work environment. An agent-based assessment model for matching between employees and tasks is presented, and two algorithms to allocate tasks to employees, namely, the minimal matching and the greedy matching algorithm, are designed. The two algorithms are then translated into the multi-agent simulation system. The simulation experiment results showed that minimal matching is better than greedy matching for rapid task allocation. Moreover, with the minimal matching algorithm, the different preference of manager has a rarer impact on completion effect of tasks than on the increase of individual capability; the following effects are evident:the different percentage of generalists and specialists have a distinct effect on task completion; the higher rate of individual capability is positively correlated with collaborative learning rate; and the different dynamic of tasks have different degree impact on the average rate of increase of capability of employees.
【Key words】 E-commerce; Collaborative work; Integrated simulation; Multi-agent based simulation; Agent evolution algorithm; Evolutionary game; Discrete simulation;

 参考文献:
[1] 龙跃,易树平.  两阶段决策下物流任务联盟协同管理优化[J]. 计算机集成制造系统. 2010(04)
[2] 谭劲松,何铮.  集群自组织的复杂网络仿真研究[J]. 管理科学学报. 2009(04)
[3] 董升平,胡斌.  人-任务互动的并行集成模拟[J]. 系统工程理论与实践. 2009(08)
[4] 陈希,樊治平,李玉花.  个体及协同因素下的员工与工作任务匹配问题[J]. 工业工程与管理. 2009(02)
[5] 吴江,胡斌,刘天印.  交互记忆系统影响人群与工作交互的模拟研究[J]. 管理科学. 2009(01)
[6] 孙虹,徐康宁.  TMT组合属性视角的决策解析——以我国企业专用性投资行为分析为例[J]. 科研管理. 2009(01)
[7] 杨阳,荣智海,李翔.  复杂网络演化博弈理论研究综述[J]. 复杂系统与复杂性科学. 2008(04)
[8] 蒋国银,胡斌.  人群工作协同进化动力学模型及模拟[J]. 科学学与科学技术管理. 2008(12)
[9] 吴婷,胡斌.  组织文化演变的定性模拟分析(英文)[J]. 系统仿真学报. 2008(19)
[10] 王艳梅,赵希男,靖可.  存在生产外部性的任务分配问题分析[J]. 管理科学. 2008(04)


  


上一篇:电子商务人才需求分析研究
下一篇:面向电子商务的[CQVIP免费论文网]谈判支持系统研究,电子商务[CQVIP免费论文网]谈文范文,电子商务[CQVIP免费论文网]谈文如何写,电子商务[CQVIP免费论文网]谈文如何写大纲

网友评论

我来说两句

用户名:   验证码: 验证码,看不清楚?请点击刷新验证码