多无人机模糊多目标分布式地面目标协同追踪

Fuzzy multi-objective distributed cooperative tracking of ground target for multiple unmanned aerial vehicles

针对城市环境中多约束条件下多无人机协同追踪地面目标问题,综合考虑具有不同重要性等级的多个优化目标,提出了一种基于分布式预测控制的模糊多目标航迹规划方法.首先,考虑城市环境中建筑物对无人机视线遮挡、无人机和传感器能量消耗等因素,分别采用目标覆盖度、控制输入代价和开关量形式传感器能耗等为目标函数,将多无人机协同追踪航迹规划转化为多目标优化问题;然后,基于分布式预测控制框架,利用每架无人机未来有限时域内的预测状态,构建多无人机之间的避碰约束,并结合最小转弯半径等约束,形成分布式协同航迹规划模型;最后,针对多个优化目标的不同重要性等级要求,利用模糊满意优化思想将目标模糊化,并根据更重要目标具有更重要满意度的原则,将优先等级表示为松弛满意度序,通过在线求解得到有限时域内每架无人机的局部航迹;与传统多目标加权算法仿真结果对比,验证了所提方法的有效性,充分说明了该方法能够获得同时满足目标优化和重要性等级要求的最优航迹.

For multiple unmanned aerial vehicle(UAV)cooperative tracking ground target with diverse constrains in urban environment,a fuzzy multi-objective path planning method based on distributed predictive control is proposed to deal with multiple objectives with different importance levels.Firstly,the factors including line of sight occlusion from buildings,energy consumptions of UAVs and sensors are considered.Correspondingly,the objective functions are designed as target coverage degree,control input cost and sensor energy consumption with switch value respectively,so that the multiple UAV cooperative tracking problem is transformed into a multi-objective optimization problem;Then,based on distributed predictive control,the predictive states of each UAV in a finite horizon are exchanged to build up collision avoidance constraint between UAVs.Combining the minimum turning radius constraints,the distributed cooperative path planning model is formulated.Finally,for the different important levels requirement,all the objectives are fuzzified by fuzzy satisfactory optimization concept.According to the principle that the objective with higher priority will have higher satisfactory degree,preemptive priorities are modeled into the relaxed order of satisfactory degrees.The local preferred path of each UAV is worked out online in a finite period.Comparing with the traditional weighted multi-objective optimization algorithm,the simulation results show the effectiveness of the proposed method.The best path satisfying the requirements of multi-objective optimization and importance levels can be obtained.