摘要
针对机动目标跟踪实际应用中所存在的机动模型不确定和非线性滤波的两大难题,文章提出了基于Elman神经网络的机动目标跟踪滤波的新算法。在常速度模型的基础上,使用Elman神经网络算法,通过对目标状态预测值与最优估计的残差、新息以及滤波增益矩阵进行在线学习,获得目标机动项的大小和过程噪声协方差矩阵的自适应调节因子,实时调整最优估计和运动模型。大量的仿真实验结果表明:本文算法有效地降低了目标运动过程中机动项对运动模型的干扰,提升了滤波性能。在强机动的条件下,滤波器的跟踪滤波性能远优于CV模型,比基于支持向量回归跟踪滤波算法模型具有更好的跟踪性能。
Aiming at the problem that the optimal allocation problem of the task route level in the route network is large and the algorithm is difficult to meet the performance requirements.This paper disassembles the problem into two steps,firstly looking for the optimal part and then micro-seeking the local optimum.The optimization effect is greatly reduced while the time overhead is greatly reduced.In the research process,the main flight height optimization method based on genetic algorithm and the segment height optimization adjustment method based on environment and individual self-balancing are proposed.The flight height of each route segment is reasonably allocated,and the altitude level of the entire route network is fully utilized.The flow capacity reduces the probability of air battlefield traffic conflicts and increases the enforceability of the plan.
作者
朱涛
Zhu Tao(The 28th Research Institute of China Electronic Technology Group Corporation,Nanjing 210007,China)
出处
《信息化研究》
2021年第4期16-20,24,共6页
INFORMATIZATION RESEARCH