基于IBES优化的三维无人机自主避障

Optimal Trajectory Planning and Autonomous Obstacle Avoidance Method for 3D UAV Based on IBES Optimization

针对三维空间环境复杂,航迹规划计算量较大,而现有BES算法受路径搜索能力不足等因素制约,无人机易在低空自主避障时陷入局部最优、难以完成复杂任务等缺陷的情况,提出了一种基于IBES算法的三维无人机自主避障方法。首先,构建威胁源模型、无人机物理约束模型以及三维山地模型,建立代价函数;其次,设计了随机Tent映射来初始化种群,提高初始化种群的质量;然后,针对BES算法在第一阶段——选择搜索空间阶段过早收敛,易陷入局部极值点的情况,引入Levy飞行策略修订此阶段更新公式,跳出局部最优;再后,设计了融合自适应指数权重的黄金正弦指引机制,提高秃鹰在既定空间全面探索并利用搜索空间的能力,解决了BES算法在既定空间内搜索猎物能力不足的问题;最后,设计了动态选择自适应t分布变异算子,提高了全局能力,同时,利用样条插值随机取点解决了路径点过于密集的问题,进一步提升了算法精度。仿真实验结果表明,提出的IBES算法规划的路径相比灰狼算法、改进的飞蛾扑火算法、麻雀算法及秃鹰搜索算法,航程分别降低了23.05、10.55、13.44和3.20 km,代价相比其他4种算法分别降低了7.26、9.51、8.27和4.91,IBES算法的寻优成功率达到了98%。同时,IBES算法能耗更优,路径平滑,收敛速度更快,使无人机有效地利用地形优势来躲避威胁源,表现出较好的寻优能力。

In view of the complexity of 3D space environment,the computation of flight path planning is large,the existing BES algorithm is restricted by the lack of path search ability and other factors,so in the low altitude autonomous obstacle avoidance,the Unmanned Aerial Vehicle(UAV)is easy to fall into the local optimum and difficult to complete complex tasks.A 3D autonomous obstacle avoidance method based on IBES is proposed.Firstly,the threat source model,UAV physical constraint model and 3D mountain model were constructed,and the cost function was established.Secondly,a random Tent mapping is designed to initialize the population and improve the quality of the initialization population.Then,aiming at the situation that BES converges too early in the first stage of selecting search space and is prone to fall into local extremal points,Levy flight strategy is introduced to revise the updating formula in this stage and get out of the local optimum.A gold sinusoidal guidance mechanism combined with adaptive index weights was designed to improve the ability of vulture needles to comprehensively explore and utilize the search space in the given space,then the problem of the insufficient ability of BES to search prey in the given space is solved.Finally,a dynamic selection adaptive t-distribution mutation operator is designed to improve the global ability.At the same time,spline interpolation is used to randomly pick points to solve the problem of too dense waypoints,which further improves the accuracy of the algorithm.Experimental simulation results show that compared with GWO,GAMFO,SSA and BES algorithms,the proposed IBES method reduces the range by 23.05,10.55,13.44 and 3.20km,and the cost is reduced by 7.26,9.51,8.27 and 4.91,respectively.The success rate is 98 percent.At the same time,IBES has better energy consumption,smooth path and faster convergence speed,so that the UAV can effectively use the terrain advantage to avoid the threat source,showing better optimization ability.