Wang Yi'an, Li Kun, Han Ying, Yan Xinxin
State Key Laboratory of Industrial Control Technology, Zhejiang University, Hangzhou 310027, China.
Faculty of Electrical and Control Engineering, Liaoning Technical University, Huludao 125105, Liaoning, China.
ISA Trans. 2022 Oct;129(Pt A):230-242. doi: 10.1016/j.isatra.2021.12.014. Epub 2021 Dec 20.
Real-time tracking of the dynamic intrusion targets consists of two crucial factors: the path forecast of the target and real-time path optimization of multi-UAV target tracking. For the first one, the uncertainty of the target trajectory is an obstacle to realizing real-time tracking. Thus a trajectory prediction method is proposed in this paper to ensure the sampling period of the target. Owing to the poor prediction accuracy of the single-step trajectory, a multi-step Unscented Kalman Filter (MUKF) is proposed to forecast its multi-step trajectory further in different regions. For the second one, there are two problems: poor optimization accuracy of the tracking trajectory and larger local optimization deviation, which will cause failure of the regional tracking. Under this circumstance, a hybrid algorithm called SAQPSO is proposed, combining the specific mechanism of two intelligence algorithms. The annealing mechanism in the Simulated Annealing (SA) algorithm is used to modify the Quantum Particle Swarm Optimization (QPSO) algorithm. Then the characteristic of quantum particles is used to update the population and enhance global searchability. Furthermore, to testify the effectiveness of the trajectory optimization algorithm and related target prediction method, a specific simulation environment is given as an example, in which the tracking trajectories of eight different algorithms are compared. Simulation results show the effectiveness of the proposed algorithm.
目标的路径预测以及多无人机目标跟踪的实时路径优化。对于第一个因素，目标轨迹的不确定性是实现实时跟踪的一个障碍。因此，本文提出一种轨迹预测方法以确保目标的采样周期。由于单步轨迹预测精度较差，提出一种多步无迹卡尔曼滤波器（MUKF）以在不同区域进一步预测其多步轨迹。对于第二个因素，存在两个问题：跟踪轨迹的优化精度差以及局部优化偏差较大，这将导致区域跟踪失败。在这种情况下，提出一种名为SAQPSO的混合算法，它结合了两种智能算法的具体机制。模拟退火（SA）算法中的退火机制用于修改量子粒子群优化（QPSO）算法。然后利用量子粒子的特性更新种群并增强全局搜索能力。此外，为验证轨迹优化算法和相关目标预测方法的有效性，给出一个具体的仿真环境作为示例，其中比较了八种不同算法的跟踪轨迹。仿真结果表明了所提算法的有效性。
