信息工程与人工智能学院知识库

  物流服务供应链在运作过程中总是存在众多不确定因素，也正是这种随机不确定性的存在，很大程度上会造成物流服务供应链供求结构失衡，因此基于需求和产出均存在随机扰动的前提，研究物流服务供应链能力决策问题非常必要。

  基于此，本文主要研究两方面的内容：第一，物流服务供应链能力决策研究，基于博弈论，构建单周期静态博弈下的物流服务供应链能力决策模型，并将单周期静态博弈下能力决策的静态解与系统动力学相结合，研究物流服务能力决策变量的多周期动态演化，通过研究发现，多周期博弈过程中会造成决策误差增大即存在牛鞭效应；第二，基于信息共享的物流服务供应链能力决策优化研究。主要就是通过构建基于信息共享的物流服务能力决策模型以期优化决策，但是在现实中，真正做到信息共享存在很大阻力，成员往往都不会轻易公开私有的信息，为了能有针对性地激励物流服务供应链节点企业选择信息共享，在本文最后对集成商和提供商的信息共享行为进行了分析。

本文的主要研究成果如下：

1. 在需求和产出均存在随机扰动的现实条件下，基于博弈论构建物流服务供应链能力决策模型，研究单周期静态博弈下的集成商和提供商的决策策略，结果表明能力投资量的值明显大于订购量的值，两个变量均受到能力产出因子、零售价等的影响，并且能力订购量与能力产出量呈正向相关关系，引入能力订购—投资储备系数来表示两者之间的关系，该系数除了受批发价、单位惩罚成本、残值、单位能力投资成本的影响外，也受产出率的影响，产出率均值越大，该系数越大，产出率方差越大，该系数越小。
2. 将博弈论与系统动力学相结合，分析能力决策值的动态演化规律。结果表明，物流服务提供商的投资量对订购量有着较高的依赖性，从物流服务能力决策变量（物流服务能力订购量和物流服务能力投资储备量）、物流服务能力持有量、牛鞭效应值三个维度进行探究，由仿真结果可以看出：提供商的物流能力投资量均值及波动范围明显大于市场需求，物流服务能力持有量变动幅度较大，这些仿真结果都证明了物流服务供应链能力决策中存在问题，牛鞭效应值达3.55，这表明基础模型的物流服务能力决策过程中产生了牛鞭效应。
3. 将信息共享机制引入到物流服务供应链能力决策基础模型中，结果表明，相比于基础模型，信息共享模型下物流能力决策量的波动幅度减小，同时物流能力供给相对较为稳定，不再需要提前准备过多的物流服务能力，便能更好地应对市场需求的变化，并且牛鞭效应值减小为1.43，在一定程度上减少了对人力和物力等资源的浪费，证明信息共享机制对于物流服务供应链能力决策有一定的协调优化作用。

  基于演化博弈理论构建物流服务供应链节点企业信息共享行为选择模型，对集成商和提供商信息共享行为的演化发展过程以及受各要素的影响进行具体分析，研究发现：双方均选择信息共享策略和双方均选择信息不共享策略是一种动态稳定的策略，而且对决策者选择信息共享意愿的行为起到反向阻碍影响的分别是信息共享成本和共享风险系数，而共享信息量、风险因子、信息吸收能力、协同系数和激励系数对其有正向激励作用，存在着改变演化路径的阈值。

In the operation of logistics service supply chain is always present many uncertain factors, it is precisely that the existence of such random uncertainty that will cause the imbalance of supply and demand structure of logistics service supply chain to a large extent. Therefore, based on the premise of stochastic disturbance of demand and output, it is necessary to study the logistics service supply chain capability decision.

Based on this, this paper mainly studies two aspects: First, the ability of logistics service supply chain decision-making research, based on game theory, building a single cycle ability of logistics service supply chain decision under static game model, and the dynamic evolution of logistics service capacity decision variables is studied by combining the static solution of capacity decision and system dynamics in a single period static game.Through the research, it is found that the bullwhip effect exists in the process of multi-cycle game, which will cause the increase of decision error. Secondly, the research of logistics service supply chain capability decision optimization based on information sharing. The main goal is to optimize the decision-making by building a decision-making model of logistics service capability based on information sharing. However, in reality, there is great resistance to information sharing, and members often do not easily disclose private information. In order to motivate logistics service supply chain node enterprises to choose information sharing, At the end of this paper, the information sharing behavior of logistics service supply chain members is analyzed.

The main research results of this paper are as follows:

（1）Under the realistic condition of stochastic disturbance of demand and output, the logistics service supply chain capability decision model is constructed based on game theory. The results show that the value of capacity investment is significantly greater than the value of order, and both variables are affected by capacity output factor, retail price, etc., and the capacity order and capacity output are positively correlated. The capacity ordering-investment reserve coefficient is introduced to represent the relationship between the two. The coefficient is not only affected by wholesale price, unit penalty cost, residual value and unit capacity investment cost, but also affected by the yield rate. The larger the mean yield rate is, the larger the coefficient is, and the larger the variance of yield rate is, the smaller the coefficient is.

（2）Combining game theory with system dynamics, analyze the dynamic evolution law of ability decision value. The results show that the amount of investment of logistics service providers has a high dependence on the amount of order. The three dimensions of logistics service capacity decision-making variables (order amount of logistics service capacity and investment and reserve amount of logistics service capacity), logistics service capacity holding and bullwhip effect value are explored. The simulation results show that: Provider of logistics capability of the investment is bigger than the average and range of market demand, logistics service capability holdings range is larger, the simulation results prove the problems existing in the logistics service capability of supply chain decisions, the bullwhip effect value of 3.55, indicating that the basic model of logistics service ability to produce the bullwhip effect in the process of decision-making.

（3）The information sharing mechanism is introduced into the logistics service supply chain ability decision-making foundation model, the results show that compared with the basic model of information sharing model volatility of logistics ability decision amount is reduced, the logistics capability to supply relatively stable at the same time, no longer need to prepare too much ahead of logistics service ability, can better respond to the change of market demand, Moreover, the bullwhip effect value is reduced to 1.43, which reduces the waste of human and material resources to a certain extent, proving that information sharing has a certain coordination and optimization effect on logistics service supply chain capability decision.

（4）Based on evolutionary game theory, a selection model of information sharing behavior of nodal enterprises in logistics service supply chain is constructed to analyze the development and evolution process of information sharing behavior of nodal enterprises and the influence of various factors. The research finds that: Both sides choose information sharing strategies and both sides not to choose information sharing strategy is a dynamic stable strategy, but also for policy makers to choose reverse in information sharing behavior affects the cost of information sharing and risk sharing coefficient, and sharing of information, risk factors, information absorption ability, coordination and incentive coefficient have a positive incentive to its, And there are thresholds for changing evolutionary paths.