首页|期刊导航|空天预警研究学报|一种拦截型无人机的硬件和算法设计

一种拦截型无人机的硬件和算法设计OA

Hardware and algorithm design of an interception-type unmanned aerial vehicle

中文摘要英文摘要

针对现有拦截型无人机应对高机动与集群目标时响应慢、效费比低、附带损伤高等问题,设计了一种拦截型无人机系统.硬件上,采用八旋翼平台、多模态感知系统、双模可切换执行机构及高算力计算平台和通信模块实现动态拦截.拦截算法上,融合交互式多模型-无迹Kalman滤波(IMM-UKF)与长短期记忆网络(LSTM)用于目标状态估计与轨迹预测,并采用基于分布式市场机制的协同任务分配(DMMBCTA)算法和深度强化学习(DRL)修正的模型预测控制(MPC)轨迹规划算法.实验结果表明,本文设计在高机动目标的拦截成功率和时效性、集群拦截能力上均有显著突破.

Aimed at the problems of slow response,low cost-effectiveness and high collateral damage when existing interception-type unmanned aerial vehicles(UAV)deal with highly maneuverable and clustered targets,an interception-type UAV system is designed.In terms of hardware,it adopts an eight-rotor platform,a multi-mod-al perception system,dual-mode switchable actuator,and high-computing-power platforms and communication modules to realize dynamic interception.For interception algorithms,the interactive multi-model-unscented Kal-man filter(IMM-UKF)and long short-term memory network(LSTM)are fused for target state estimation and tra-jectory prediction.In addition,the distributed market mechanism-based cooperative task allocation(DMMBCTA)algorithm and the model predictive control(MPC)trajectory planning algorithm updated by deep reinforcement learning(DRL)are adopted.Experimental results show that the design in the paper has made significant break-throughs in the interception success rate,timeliness,and cluster confrontation capability of highly maneuverable targets.

陈立军

广州软件学院 软件与人工智能学院,广州 510990

信息技术与安全科学

拦截型无人机多模态感知双模可切换目标状态估计与轨迹预测协同任务分配

interception-type UAVmulti-modal perceptiondual-mode switchabletarget state estimation and trajectory predictioncooperative task allocation

《空天预警研究学报》 2026 (4)

281-286,6

广州软件学院科研项目(KY202504)

10.3969/j.issn.2097-180X.2026.04.009

评论