基于协同进化算法的导弹与飞机追逃对策

研究了导弹与飞机在复杂动力学约束下的三维追逃对策.阐述了协同进化的理论基础,根据追逃双方运动的数学模型设计了混合追逃策略以保证进化过程的推进.在简单适应度和共享适应度的基础上提出了混合适应度,利用协同进化算法对三维环境下导弹和飞机的追逃策略进行进化.仿真结果表明,进化的飞机能有效规避比例制导策略的导弹,并且在进化过程中导弹和飞机的策略表现出一定程度的改善.

Pursuit-evasion game between missile and airplane based on co-evolutionary algorithm

Under complex dynamic constraints,a three-dimensional pursuit-evasion game between missile and airplane is studied.Based on the co-evolutionary theory and the mathematical models of missile and airplane dynamics,mixed pursuit-evasion strategies are designed for both players to guarantee the evolution process.In accordance with the simple fitness and the sharing fitness,a mixed fitness is developed to evaluate game results.Then a co-evolutionary algorithm is applied to evolve the strategies of both players concurrently.The influences of different fitness on evolution results are compared.Simulation results show that the evolved airplane can effectively evade the proportionally guided pursuers.And in the co-evolutionary process,the strategies of both players are improved.