基于鲁棒优化的无人机空战博弈决策  

Game decision making of UAV aerial combat based on robust optimization

作　　者：王昱[1,2] 章卫国[1] 傅莉[2] 黄得刚 李勇[3] 

机构地区：[1]西北工业大学自动化学院,陕西西安710129 [2]沈阳航空航天大学自动化学院,辽宁沈阳110136 [3]沈阳工业大学电气工程学院,辽宁沈阳110870

出　　处：《系统工程与电子技术》2015年第11期2531-2535,共5页Systems Engineering and Electronics

基　　金：国家自然科学基金(61374032);辽宁省教育厅科学研究一般项目(L2015412);沈阳市科技创新团队(src201204)资助课题

摘　　要：作为无人机空战过程中较大的不确定性因素,作战双方毁伤概率易受外界因素影响,可能在一定范围波动,针对这一问题将鲁棒优化思想引入无人机空战博弈模型中,结合纳什均衡求解的适应度函数的期望和方差,建立了变权重的自适应目标函数,设计了权重的分配方法。在确保策略解可行的同时,提高了毁伤概率变化因素下混合策略的鲁棒性。仿真实验结果表明,与原博弈模型相比,新模型纳什均衡解下的双方博弈收益受参数变化影响较小,可提供更可靠的策略参考。As one of the large uncertainty factors in the process of unmanned aerial vehicle （UAV） aerial combat, the damage probability of both sides may fluctuate in a certain range for it is easily affected by external factors. To solve this problem, the robust optimization theory is imported into the game model of UAV aerial combat. Integrated with expectation and variance for the fitness function of Nash equilibrium solution, an adap- tive objective function with variable weight is established, and the method for weight distribution is designed. The strategy solution is feasible, meanwhile the robustness of hybrid strategies is improved in the case of the damage probability changing. The simulation results show that compared with the original game model, the game revenue for both sides obtained from Nash equilibrium solution of the new model is less affected by the change of parameters. Therefore it can provide more reliable strategies reference.

关 键 词：无人机 不确定性因素 鲁棒优化 纳什均衡 自适应权重 

分 类 号：V247.1[航空宇航科学与技术—飞行器设计] TP301.6[自动化与计算机技术—计算机系统结构]