具有学习效应的寡头厂商重复博弈均衡分析

An Equilibrium Analysis of Repeated Game Model of Oligopoly Firm with Learning Effect

讨论一个有三位厂商参加的具有学习效应的重复博弈模型,厂商们在每一阶段决定各自产量和价格,使得累积利润最大化.在市场需求函数为非线性的条件下,引进带有学习曲线的成本函数,使得重复博弈模型更切合实际.接着把重复博弈转化成求解多阶段非线性规划问题,利用牛顿法进行求解全局垄断解;在阶段最优的基础上,求得参与者的阶段最优解;进而求得精炼子博弈纳什均衡解.最后比较分析这三种解和实验结果.

A repeated game model of oligopoly firm with learning effect is discussed. In each stage, the three players should decide product and price in order to maximize the cumulated profit. Under the condition of nonlinear demand function with the cost function defined by learning curve, the repeated game as a multistage dynamic programming problem is formulated, and solve the global monopoly optimal solution, local (stage) optimal solution, and finally the subgame perfect Nash equilibrium. At last, the comparisons among the three kinds of solutions and the experiment solution were made.