中文摘要：

空间博弈研究中多假定网络上物理相邻的个体必然发生博弈，忽略了真实系统中个体与邻居发生交互不具确定性。从合作演化的角度，利用空间囚徒博弈理论，依据策略性质对博弈分类，提出一种“随机交互强度”机制，使博弈以固定概率随机地发生在相邻个体间，建立了基于二维正方晶格网络的博弈演化模型，并使用多主体系统进行了仿真。利用系统平均合作水平作为判据，对比连续策略系统与离散策略系统的不同交互强度，给出了仿真结果。结果表明，个体间博弈的随机性可以提高离散策略系统的平均合作水平，较低的交互强度可使系统达到合作稳定态，相比于必然性交互机制，上述机制可以有效促进持离散策略个体间的合作。

英文摘要：

Most researches of evolution of cooperation assume that two individuals interact with each other if they are physical neighbors. However,an individual is involved in the game with a neighbor in a certain probability ac- cording to the environment and risk tendency in the real social system. From the perspective of the evolution of coop- eration ,interaction stochastieity mechanism is proposed based on spatial prisoner＇s game theory and game strategy classification. Game takes place between the adjacent individuals randomly with a fixed probability. A evolutionary game model based on two - dimensional lattice network was established. The simulation system was built up based on multi - agent system. The average cooperation level of the system was used as a criterion and the continuous strategy system and the discrete strategy system were compared. Simulation results show that the low interaction rate can ef- fectively promote the cooperation level in discrete strategy system compared with inevitable interaction mechanism.