针对多无人机在无源定位过程中协同动态规划航迹提高定位精度问题，提出基于演化深度神经网络的分布式动态航迹优化方法。首先将演化计算与深层前向反馈神经网络结合，设计基于演化神经网络的无人机协同无源定位动态航迹规划框架。以多无人机到达角（AOA）协同定位为例，利用定位过程中对目标估计的克拉美罗界（CRLB）生成最优训练集。通过无人机下一时刻与目标形成的相对构型作为系统学习的行为，从而得到下一时刻优化后的航迹点。实验结果表明，该方法相对于传统中心控制的无人机协同定位方法，具有更低的处理延时，能够以更短时间达到定位精度。
Aiming at the path planning problem of multiple unmanned aerial vehicles (UAVs) in passive localization, an unmanned aerial vehicle dynamic path planning method based on evolutionary depth neural network is proposed. Firstly, this method combines the differential evolution algorithm and BP neural network, and designs a learning path planning framework for UAV passive location based on evolutionary neural network. Then, angle of arrival (AOA) localization is used for the multiple UAVs, and an optimal training set is generated based on the Cramer-Rao low bound (CRLB) of target estimation. The optimized waypoints can be acquired from the learning behavior of the relative deployment between UAVs and target. Experimental results show that the unmanned aerial vehicle (UAV) based on the evolutionary neural network can greatly improve real-time performance and decrease location time.
[1] Li X, Deng Z D, Rauchenstein L T, et al. Contributed review:Source-localization algorithms and applications using time of arrival and time difference of arrival measurements[J]. Review of Scientific Instruments, 2016, 87(4):41-50.
[2] Adelipour S, Hamdollahzadeh M, Behnia F. Constrained optimization of sensors trajectories for moving source localization using tdoa and fdoa measurements[C]//Proceedings of the 3rd RSI International Conference on Robotics and Mechatronics. Tehran, Iran:Robotics Society of Iran, 2015:200-204.
[3] Xu S, Dogancay K, Hmam H. Distributed pseudolinear estimation and UAV path optimization for 3D AOA target tracking[J]. Signal Processing, 2017(133):64-78.
[4] Kruusmaa M. Global level path planning for mobile robots in dynamic environments[J]. Journal of Intelligent and Robotic Systems, 2003, 38(3):55-83.
[5] Chen C, Li H, Dong D. Hybrid control for robot navigation-a hierarchical q-learning algorithm[J]. IEEE Robotics & Automation Magazine, 2008, 15(2):37-47.
[6] Zhao Y, Li W, Shi P. A real-time collision avoidance learning system for unmanned surface vessels[J]. Neurocomputing, 2016(182):255-266.
[7] Zhang B, Mao Z, Liu W. Optimality analysis of sensor-target localization geometries[J]. Journal of Intelligent & Robotic Systems, 2015, 77(2):391-409.
[8] Xu S, Dogancay K. Optimal sensor placement for 3D angle-ofarrival target localization[J]. IEEE Transactions on Aerospace and Electronic Systems, 2017, 53(3):1196-1211.
[9] Wang W, Bai P, Liang X. Performance analysis for TDOA localization using UAVs with flight disturbances[C]//Proceedings of the 20th International Conference on Information Fusion. Xi'an:IEEE, 2017.
[10] Dogancay K. UAV path planning for passive emitter localization[J]. IEEE Transactions on Aerospace and Electronic Systems, 2012, 48(2):1150-1166.
[11] Cordón O. A historical review of evolutionary learning methods for Mamdani-type fuzzy rule-based systems:Designing interpretable genetic fuzzy systems[J]. International Journal of Approximate Reasoning, 2011, 52(6):894-913.
