基于遗传算法的UCAV动态任务分配模型及研究

Dynamic UCAV Mission Assignment Using Genetic Algorithm

针对多UCAV(multi-Unmanned Combat Aerial Vehicle)动态任务规划问题提出了单任务类型动态规划模型,基于事件驱动的原则,满足对打击不成功目标和新出现目标需要进行动态规划的要求,在实施目标攻击后或新目标出现时,以获得最大收益值为目标,采用遗传算法进行动态目标分配。与多任务类型模型相比,单任务类型模型有效地减小了可能解规模,寻优过程计算量小,所需时间短,仿真实验验证了模型的有效性。

A single mission model was used in multi-UCAV （Unmanned Combat Aerial Vehicle） dynamic mission assignment problem. The event-driven rule takes unsuccessful attack and new targets＇ appearance into consideration, after targets attack or new targets appear, using genetic algorithm to assign UCAV to mission on the maximize of expectation value. Comparing to multi-task model, single task model can effectively reduce the size of likely solutions, which will lead to shorter calculation time while using Genetic Algorithm. Simulation result shows the single task model has great superiority in large scale dynamic mission assignment situations.