Abstract: Decision-making model of speeding and no speeding of drivers and inspection and no inspection of traffic mangers is proposed aiming at the speeding behavior of drivers. The evolutionary process that traffic mangers and drivers choose strategies is analyzed using evolutionary game. The possible evolutionary stable states are analyzed accordingly to local stability of Jacobian matrix. The evolutionary stable state has several kind conditions such as saddle point, stability point and instability point because of traffic mangers and drivers’ different benefit. The results show that government department must reward traffic mangers for doing their duty and punish traffic mangers for their breach of duty to urge traffic mangers to inspect strictly. At the same time, the speeding drivers should be punished strictly according to individual difference. Thus speeding driving can be forbidden to a certain extent and improve traffic safety.

Key words: traffic engineering, speeding behavior, evolutionary game, stable state

摘要： 为遏制超速驾驶，通过分析驾驶员超速的心理博弈过程，建立了交通管理者和驾驶员策略选择的决策博弈模型，应用演化博弈论分析了交通管理者检查与不检查、驾驶员超速与不超速各自2种策略选择的时间演化过程. 根据雅可比矩阵的局部稳定性对可能的演化稳定状态进行了分析，由于交通管理者和驾驶员选择不同的策略会有不同的收益，因此，演化稳定状态会有鞍点、稳定点、不稳定点等几种情况. 分析结果表明：政府部门要加强对交通管理者尽职的奖励与失职的惩罚，敦促交通管理者严格检查；同时，针对个体差异对超速的驾驶员进行严厉处罚，这样就会在一定程度上遏制超速驾驶，提高交通安全.

关键词: 交通工程, 超速, 演化博弈, 稳定状态