| 引用本文: | 赵玉亮,宋业新,张建军,康丽文.基于多策略融合粒子群的无人机对地攻击模糊博弈决策[J].控制理论与应用,2019,36(10):1644~1652.[点击复制] |
| zhaoyuliang,SONG Ye-xin,ZHANG Jian-jun,KANG Li-wen.Fuzzy game decision-making of unmanned aerial vehicles air-to-ground attack based on the particle swarm optimization integrating multiply strategies[J].Control Theory and Technology,2019,36(10):1644~1652.[点击复制] |
|
| 基于多策略融合粒子群的无人机对地攻击模糊博弈决策 |
| Fuzzy game decision-making of unmanned aerial vehicles air-to-ground attack based on the particle swarm optimization integrating multiply strategies |
| 摘要点击 2268 全文点击 995 投稿时间：2018-06-13 修订日期：2019-05-18 |
| 查看全文 查看/发表评论 下载PDF阅读器 |
| DOI编号 10.7641/CTA.2019.80437 |
| 2019,36(10):1644-1652 |
| 中文关键词 无人机 协同对地攻击 不确定性 博弈论 可能性理论 熵权法 粒子群 |
| 英文关键词 UAVs cooperative air-to-ground attack uncertainty game theory necessity theory entropy weight method particle swarm optimization |
| 基金项目 国家自然科学基金 |
|
| 中文摘要 |
| 针对无人机协同对地攻击的复杂性和不确定性，联合防空火力压制与对地目标打击任务，引入存活因子、摩擦因子和状态因子等概念，考虑目标威胁度的模糊性，结合生存概率和武器消耗等因素，建立一种多阶段的模糊多目标任务分配规划模型.为更好地描述攻击任务的对抗性和多策略性，以博弈论为框架，将规划模型转化为模糊多目标双矩阵博弈综合集结模型；利用必要性理论将集结模型中的不确定性目标清晰化处理，进而运用熵权法对多个目标进行加权求和，将其转化为单目标双矩阵博弈模型.提出基于多策略融合粒子群算法的纳什均衡求解方法，通过引入自适应惯性权重、动态反向学习与局部变异策略，在增强种群多样性的同时保证粒子群局部精确搜索能力.算例仿真结果验证了所提模型和方法的有效性. |
| 英文摘要 |
| In view of the complexity and uncertainty of cooperative air-to-ground attack for unmanned air vehicles (UAVs), a multi-stages and fuzzy multi-objective programming is proposed by jointing the mission of suppression of enemy air defense and the mission of ground target attack, introducing the concepts of survival factor, friction factor and state factor and combining the factors of survival probability, weapon consumption and fuzzy target threat. In order to better describe the antagonism and multi-strategic nature of the attack task, the game theory is used to transform the planning model into a synthetic aggregation model of fuzzy multi-objective bi-matrix game. Using the theory of necessity, the uncertainty objective in the aggregation model is clarified. And then using the entropy weight method, the aggregation model can be solved by transforming it into a single-target bi-matrix game model. A method of solving Nash equilibrium based on the particle swarm optimization algorithm integrating multiply strategies is proposed. By introducing adaptive inertia weight, dynamic inverse learning and local mutation strategy, the local precise search capability of particle swarms can be guaranteed while enhancing population diversity. The simulation results of the example verify the validity of the proposed model and method. |