摘要
针对无人机对峙跟踪非广域目标问题,开展目标状态估计与无人机制导方法研究。首先建立非广域地理环境模型,将非广域地理约束作为伪观测方程引入粒子滤波器的观测方程。其次,鉴于目标在运动过程中可能受到多个模型的约束,采用交互多模型滤波算法进行状态估计,即每个模型对应的受约束粒子滤波器并行工作,并对多个滤波器估计结果进行加权,得到更精确的目标运动状态估计值。然后,提出时间最优导航向量场,通过计算期望航向角,引导无人机快速收敛至目标极限环。最后,仿真实验表明,受约束粒子滤波-多交互模型算法相比于传统的滤波算法,估计精度提高了20%,时间最优导航向量场方法相比于传统的导航向量场方法,引导效率提高了15%,所提方法可更有效地用于解决非广域目标对峙跟踪问题。
In this paper,the research of unmanned aerial vehicle(UAV)standoff tracking for target in non-wide area is conducted.Firstly,the geographical environment model of non-wide area is established,and the geographical constraint of non-wide area is introduced as a pseudo observation equation into the observation equation of the particle filter(PF).Secondly,given that a target may be constrained by multiple models,the interacting multiple model(IMM)filter algorithm is used for state estimation.That is,the constrained particle filter(CPF)corresponding to each model works in parallel,and the estimation results of multiple filters are then weighted to obtain a more accurate estimation result of the target's motion state.Then,the time-optimal guidance vector field(TOGVF)is proposed.By calculating the desired heading angle,it guides the UAV to quickly converge to the limit cycle above target.Finally,simulation experiments show that the CPF-IMM algorithm improves the estimation accuracy by 20%compared with the traditional filtering algorithm,and the TOGVF method improves the guidance efficiency by 15%compared with the traditional guidance vector field method.
作者
姚鹏
钟晨
YAO Peng;ZHONG Chen(College of Engineering,Ocean University of China,Qingdao 266100,China)
出处
《无人系统技术》
2024年第1期78-86,共9页
Unmanned Systems Technology
基金
国家自然科学基金(51909252)
山东省自然科学基金(ZR2023ME009)。
关键词
无人机
对峙跟踪
非广域目标
伪观测方程
受约束粒子滤波
交互多模型
时间最优导航向量场
Unmanned Aerial Vehicle
Standoff Tracking
Target in Non-wide Area
Pseudo Observation Equation
Constrained Particle Filter
Interacting Multiple Model
Time-optimal Guidance Vector Field