The jamming resource allocation problem of the aircraft formation cooperatively jamming netted radar system is investigated.An adaptive allocation strategy based on dynamic adaptive discrete cuckoo search algorithm(DA...The jamming resource allocation problem of the aircraft formation cooperatively jamming netted radar system is investigated.An adaptive allocation strategy based on dynamic adaptive discrete cuckoo search algorithm(DADCS)is proposed,whose core is to adjust allocation scheme of limited jamming resource of aircraft formation in real time to maintain the best jamming effectiveness against netted radar system.Firstly,considering the information fusion rules and different working modes of the netted radar system,a two-factor jamming effectiveness evaluation function is constructed,detection probability and aiming probability are adopted to characterize jamming effectiveness against netted radar system in searching and tracking mode,respectively.Then a nonconvex optimization model for cooperatively jamming netted radar system is established.Finally,a dynamic adaptive discrete cuckoo search algorithm(DADCS)is constructed by improving path update strategies and introducing a global learning mechanism,and a three-step solution method is proposed subsequently.Simulation results are provided to demonstrate the advantages of the proposed optimization strategy and the effectiveness of the improved algorithm.展开更多
为解决快速扩展随机树算法(rapid-exploration random tree,RRT*)在三维环境中盲目搜索路径以及缺乏节点扩展记忆性等问题,提出一种融合蚁群算法的双向搜索算法ACO-RRT*。为适应精细化三维建模环境和解决地面起伏不平坦等问题,对RRT*算...为解决快速扩展随机树算法(rapid-exploration random tree,RRT*)在三维环境中盲目搜索路径以及缺乏节点扩展记忆性等问题,提出一种融合蚁群算法的双向搜索算法ACO-RRT*。为适应精细化三维建模环境和解决地面起伏不平坦等问题,对RRT*算法进行改进优化。采用双向搜索策略,在起点和终点同时运行改进后的RRT算法和蚁群算法,相向而行,对路径长度和运行时间进行优化。针对生成路径不够平滑等问题,引入B样条曲线平滑策略优化路径。仿真结果表明,所提算法能够有效用于机器人三维路径规划。展开更多
为解决清漂船在复杂水面背景下对漂浮物体积较小或被遮挡的检测与跟踪问题,提出了一种基于视觉的水面背景下目标检测与跟踪算法,通过利用YOLO改进的多粒度特征融合方法使得模型在最终检测时所提取的特征向量考虑更多底层的特征,并引入K...为解决清漂船在复杂水面背景下对漂浮物体积较小或被遮挡的检测与跟踪问题,提出了一种基于视觉的水面背景下目标检测与跟踪算法,通过利用YOLO改进的多粒度特征融合方法使得模型在最终检测时所提取的特征向量考虑更多底层的特征,并引入K邻域搜索感兴趣区域模块,与长短期记忆神经网络(long-short term memory,LSTM)相结合,弥补了卷积神经网络的时序关联性差的缺陷,根据目标当前帧语义特征和运动特征预测下一帧中目标所在位置,能够更快地提取目标的特征,并且有效地去除复杂背景的干扰。实验结果表明:该算法的跟踪平均成功率、平均准确率和平均速度分别为57.1%、71.1%、45.4帧/s。较好解决因被检测目标过小的问题,提升在跟踪目标被遮挡的情况下的跟踪性能。展开更多
文摘The jamming resource allocation problem of the aircraft formation cooperatively jamming netted radar system is investigated.An adaptive allocation strategy based on dynamic adaptive discrete cuckoo search algorithm(DADCS)is proposed,whose core is to adjust allocation scheme of limited jamming resource of aircraft formation in real time to maintain the best jamming effectiveness against netted radar system.Firstly,considering the information fusion rules and different working modes of the netted radar system,a two-factor jamming effectiveness evaluation function is constructed,detection probability and aiming probability are adopted to characterize jamming effectiveness against netted radar system in searching and tracking mode,respectively.Then a nonconvex optimization model for cooperatively jamming netted radar system is established.Finally,a dynamic adaptive discrete cuckoo search algorithm(DADCS)is constructed by improving path update strategies and introducing a global learning mechanism,and a three-step solution method is proposed subsequently.Simulation results are provided to demonstrate the advantages of the proposed optimization strategy and the effectiveness of the improved algorithm.
文摘为解决快速扩展随机树算法(rapid-exploration random tree,RRT*)在三维环境中盲目搜索路径以及缺乏节点扩展记忆性等问题,提出一种融合蚁群算法的双向搜索算法ACO-RRT*。为适应精细化三维建模环境和解决地面起伏不平坦等问题,对RRT*算法进行改进优化。采用双向搜索策略,在起点和终点同时运行改进后的RRT算法和蚁群算法,相向而行,对路径长度和运行时间进行优化。针对生成路径不够平滑等问题,引入B样条曲线平滑策略优化路径。仿真结果表明,所提算法能够有效用于机器人三维路径规划。
文摘为解决清漂船在复杂水面背景下对漂浮物体积较小或被遮挡的检测与跟踪问题,提出了一种基于视觉的水面背景下目标检测与跟踪算法,通过利用YOLO改进的多粒度特征融合方法使得模型在最终检测时所提取的特征向量考虑更多底层的特征,并引入K邻域搜索感兴趣区域模块,与长短期记忆神经网络(long-short term memory,LSTM)相结合,弥补了卷积神经网络的时序关联性差的缺陷,根据目标当前帧语义特征和运动特征预测下一帧中目标所在位置,能够更快地提取目标的特征,并且有效地去除复杂背景的干扰。实验结果表明:该算法的跟踪平均成功率、平均准确率和平均速度分别为57.1%、71.1%、45.4帧/s。较好解决因被检测目标过小的问题,提升在跟踪目标被遮挡的情况下的跟踪性能。
