Abstract: This paper, combining the characteristics of geographically distributed information systems development team with the study of the evolution of trust, builds the theory model of GDISD team’s trust evolution. And from the view of the evolutionary game theory to describe the mechanism of trust evolution between any two teams in GDISD team members, then verify the premise of trust evolution theory model. Study shows that the theoretical model of the trust evolution in GDISD team：the based calculation trust, rapid trust, understanding-based trust and evolve to the identification-based trust. In view of the GDISD process, analyze the foundation and object of the theoretical model about in various stages of trust development. By building dynamic replication equation, system equilibrium points are derived. And Jacobian matrix is adopted to verify their stability. The results show that when the game players both take the high-trust policy, bonus obtaining from the efficient completion of the project greater is than the loss of revenue due to the game of one other adopting the low-trust policy, the evolution of the trust all follow the evolution path from a low-trust policy to high-trust policy.

Key words: geographically distributed information systems development, team trust, evolutionary game, Jacobian matrix

摘要： 结合异地分布式信息系统开发（GDISD）团队的特点与信任演进的相关研究构建GDISD团队的信任演进理论模型。从演化博弈论的角度，描述GDISD团队任意两成员间信任演进的机理，验证信任演进理论模型成立的前提。研究得出GDISD团队信任演进的理论模型，从基于计算的信任，快速信任，基于了解的信任演化到基于认同的信任。针对GDISD过程，分析了该理论模型信任发展各阶段信任的基础和对象。通过构建动态复制方程，得出系统均衡点，并采用雅可比矩阵验证其稳定性，结果显示当博弈双方采取高信任策略时，项目高效完成的奖金大于由于博弈中一方低信任策略造成的另一方的收益损失时，信任的演化都遵循低信任策略到高信任策略的演化路径。

关键词: 异地分布式信息系统开发, 团队信任, 演化博弈, 雅可比矩阵