集成超网络和突变论的移动商务价值链孵化模拟研究

Mobile value chain incubation has become the research hotspot, due to its promotion on mobile value chain by optimizing the incubation environment. However, the value flows in the incubation system are inevitably influenced by a varity of uncertain factors, resulting in unstabitily and catastrophe phenomenon. Value chain incubation should avoid the Catastrophe to pursue its continuous growth. To study user behavior driven mobile value chain incubation mechanism and catastrophe mechanism, an integration simulation framework is proposed. Firstly, an incubation network model is built based on super-network theory, and then multi-agent simulation and discrete event simulation are intergrated to built a process-oritented simulation model to explore incubation factors and incubation mechanism. Secondly, a new paradigm combining the qualitative simulation, fuzzy function and stochastic cusp catastrophe theory is proposed to investigate catastrophe mechanism. By transforming the static and strict catastrophe mathematic model into qualitative simulaiton model, influencing factors and variable range resulting in sudden chang are analyzed accurately. Finally, Bayesian probabilistic reasoning network is introduced to built a simulation optimization model to solve multi-objective optimization problem on value chain incubation with uncertainty and group behavior. A new algorithm is proposed to maximum incubation benefit and users’ utility. Then an optimization engine with the function of simulation and optimization is designed to deal with the optimization problem. The study will contribute to management system simulation, organization theory, super-network theory.

移动商务孵化平台通过优化在孵环境，培育价值链持续成长，成为当前研究热点。然而，价值链孵化受不确定因素的影响，容易产生不稳定及突变现象。为了研究价值链孵化机理与突变规律：本课题基于超网络理论构建孵化网络模型，集成多Agent模拟和离散事件仿真，建立面向流程的价值链孵化仿真模型，研究价值链孵化的影响因素及孵化机理；集成定性模拟、模糊函数和随机尖点突变理论，将静态严格的突变数学模型转化为动态模糊的定性突变模拟模型，研究价值链孵化的突变规律，分析导致突变的因素及活动范围，精确把握预警区、临界区和突变区；引入基于实际数据的贝叶斯网络推理，构建突变约束的价值链孵化动态仿真模型，设计不确定性、群体行为化的模拟-优化方法，并开发具有模拟和优化功能的优化引擎，实现远离突变的价值链孵化的实时优化决策。本课题对管理系统模拟、超网络、突变论的集成研究具有理论意义；对提升价值链孵化的稳定性及效率具有现实意义。

在用户行为、价值链运作实时变化、外来攻击随时出现的移动互联网时代，培育移动商务价值链持续稳定成长，显得尤为重要。本项目针对移动商务价值链孵化机理与突变规律，完成以下研究：（1）构建价值链孵化超网络仿真模型。基于演化博弈理论，集成离散事件模拟和多Agent模拟，研究价值链孵化机理，为提升价值链孵化效率提供理论参考。（2）揭示价值链孵化的突变规律及控制策略。集成定性模拟、模糊函数和定性模拟，构建定性突变模拟模型，揭示价值链孵化突变规律，把握导致突变发生的因素及活动范围。从而提前规避突变的发生及级联扩散，为价值链孵化突变预防策略制定提供理论依据。（3）提出突变约束下的价值链孵化优化方法，即不确定的、群体行为化的仿真优化方法。引入基于实际数据的贝叶斯网络，集成启发式算法和仿真模型设计模拟-优化方法，研究远离突变的价值链实时最优策略。本项目研究对于丰富计算组织理论、管理系统模拟、行为运作管理以及优化理论的发展，都有一定学术意义。