基于贝叶斯估计和群体智能的无人机轨迹优化-《计算机技术与发展》

文章信息/Info

Title:: UAV Trajectory Optimization Based on Bayesian Estimation and Swarm Intelligence

作者:: 丁汝妍1; 李欢1; 莫欣岳1*; 吴灿2; 李昕雨1; 1. 海南大学网络空间安全学院(密码学院),海南海口 570228;2. 海南大学信息与通信工程学院,海南海口 570228

Author(s):: DING Ru-yan1; LI Huan1; MO Xin-yue1*; WU Can2; LI Xin-yu1; 1. School of Cyberspace Security (School of Cryptology),Hainan University,Haikou 570228,China;2. School of Information and Communication Engineering,Hainan University,Haikou 570228,China

Keywords:: unmanned aerial vehicles trajectory optimization; passive localization; Bayesian optimization; particle swarm algorithm; simulated annealing algorithm

摘要:: 为提高无人机的定位精度与队形调整效率,提出了基于贝叶斯估计的定位模型和基于群体智能算法的队形调整方法。首先,考虑实际情况中的测量噪声影响,在定弦定角模型中引入贝叶斯最大后验概率得到新的定位模型。然后,针对粒子群算法易陷入局部最优的问题,结合模拟退火算法提出改进的队形调整算法。仿真结果表明:提出的定位模型对圆形(锥形)编队的误差率比初始模型降低 72. 8% (49. 2% );改进的队形调整算法对圆形(锥形)编队的误差率相对于原始算法和遗传算法与高斯伪谱法相嵌套的方法分别降低了 37. 1% (27. 0% )和 24. 7% (19. 9% ),收敛迭代次数分别降低了 12. 5% (20% )与 12. 5% (4. 8% )。实验结果验证了提出的优化方案具有较高的精度和计算效率。

Abstract:: To improve the localization accuracy and formation adjustment efficiency of UAVs,a positioning model based on Bayesian esti-mation and a formation adjustment method based on a swarm intelligence algorithm is proposed. Firstly,considering the influence of measurement noise in the actual situation,a new localization model is obtained by introducing maximum a posteriori estimation into the fixed string fixed angle model. Then,for the problem that the particle swarm algorithm tends to fall into local optimization,an improved queue adjustment algorithm is proposed in combination with a simulated annealing algorithm. The simulation results show that the error rate of the proposed localization model for circular (conical) formation is 72. 8% (49. 2% ) lower than that of the initial model. The error rate of the improved formation adjustment algorithm for circular (conical) formation is 37. 1% (27. 0% ) and 24. 7% (19. 9% ) lower than that of the original algorithm and the combined method genetic algorithm and Gauss pseudo spectral method respectively,while the number of convergence iterations decreases by 12. 5% (20% ) and 12. 5% (4. 8% ) respectively. The experimental results verify the proposed optimization scheme’s high accuracy and computational efficiency.