In this paper, we present a trajectory generation method of a quadrotor, based on the optimal smoothing B-spline, for tracking a moving target with consideration of relative tracking pattern or limited field of view of the onboard sensor in cluttered environments. Compared to existing methods, safe flying zone, vehicle physical limits, and smoothness are fully considered to guarantee flight safety, kinodynamic feasibility, and tracking performance. To tackle the cluttered environments, a parallel particle swarm optimization algorithm is applied to find the feasible waypoints that the generated trajectory should be as close to as possible, with consideration of the target’s future state as well as obstacles to trade off the tracking performance and flight safety. Then, a sequential motion planning method, considering the above constraints, is applied and embedded into a cost function for solving the problem of robust tracking trajectory generation for the quadrotor via a convex optimization approach. The feasibility and effectiveness of the proposed method are verified by numerical simulations.

在本文中，我们提出了一种基于最佳平滑B样条的四旋翼轨迹生成方法，该方法用于在杂乱环境中考虑相对跟踪模式或车载传感器的有限视场来跟踪运动目标。与现有方法相比，充分考虑了安全飞行区域，车辆物理极限和平滑度，以确保飞行安全，运动学可行性和跟踪性能。为了解决混乱的环境，应用了并行粒子群优化算法，在考虑目标的未来状态以及权衡跟踪性能和飞行障碍的情况下，找到了生成的轨迹应尽可能接近的可行航路点。安全。然后，考虑到上述约束，采用顺序运动计划方法，应用并嵌入到成本函数中，以通过凸优化方法解决四旋翼的鲁棒跟踪轨迹生成问题。数值仿真验证了该方法的可行性和有效性。