基于惩罚机制的产学研合作演化博弈分析

The industry technology innovation alliance based cooperation by using evolutionary game theory

摘要:

产学研合作对构建国家创新体系和实现向创新驱动型经济体转变具有重要的作用,而在新形势下,多方参与的产学研合作模式成为主流,为了保证各方的利益及合作的顺利进行,在长期的合作中需要有一定的约束惩罚机制。因此本文首先建立多人多策略的演化博弈模型,并对各情况下的收益进行分析计算；然后建立惩罚机制,并把惩罚和收益关联起来,调整各方不同情况下的收益；之后使用粒子群优化算法进行模拟仿真,摒弃传统固定好数值的静态模拟方法,考虑“支出-收益”比例,“惩罚-收益”比例的动态数值分析,模拟得出惩罚约束存在下,多人多策略的产学研合作演化情况；最后,得出结论并提出相应的对策建议。

关键词:产学研合作,演化博弈,雪堆模型,惩罚,粒子群优化算法

Abstract:

Industrial technology innovation alliance can promote the industry-university-institute cooperation. Our paper established n-player and n-choice snowdrift game model to evolve cooperation. The punishment mechanisms are built up for cooperation. By simulation experiment, it can be seen that cooperation is closely relative to the benefits and, negative correlation with cost; Under the punishment mechanism, can inspire more cooperation and cooperation related to the magnitude of the punishment is, but no matter under what kind of performance mechanism, the level of cooperation and participate in the number of members in the league is in verse proportional to; Finally, we proposed some methods to improve the industry technology innovation alliance cooperation.

Key words:Cooperation; Evolutionary game theory; ISD (Iterated Snowdrift dilemma); Punishment