机场净空区非合作无人机目标监视的关键技术研究

Unauthorized operations of unmanned aerial vehicles (UAV) near airports cause great threats to the civil aviation safety. The wide range and complex environment of the airport clearance area, as well as the low, slow, and small characteristics of UAV, constitute serious challenges to the existing detection measures. In order to monitor the non-cooperative UAV targets in airport clearance area, we focus on construction method of surveillance network guided by the multi-living agent theory, collaborative detection technology using multiple sensors, and synthetic recognition technology via multiple sensors. An experimental system will be developed to verify the deployment plan, collaborative strategy, and synthetic recognition technology in the airport clearance area. This project aims to break through the key technologies such as optimal deployment of the surveillance network based on multi-objective optimization algorithm, signal processing of weak targets in fractional Fourier domain, target detection based on multi-node collaborative mechanism, and target synthetic recognition based on deep forest. The research outcomes can provide theoretical guidance for the development and application of surveillance system, which will improve the detection and recognition capabilities of non-cooperative UAV targets in the airport.

无人机黑飞对机场航班飞行安全造成了极大的威胁。机场净空区范围广、环境复杂，无人机目标具有低、慢、小等特点，对现有探测手段造成了严峻的挑战。为监视机场净空区非合作无人机目标，本项目重点研究多活性代理理论指导下的监视网络构建方法、多元传感器协同探测技术、基于多传感器的综合识别技术，并研制实验系统，在机场净空区对部署方案、协同策略和综合识别技术进行验证。本项目拟突破基于多目标优化算法的监视网络优化部署、微弱目标的分数域信号处理、基于多结点协同机制的目标检测、基于深度森林的目标综合识别等关键技术。研究成果可为监视系统的研制和应用提供理论指导，将提升机场对非合作无人机目标的探测和识别能力。

无人机黑飞对机场航班飞行安全造成了极大的威胁。机场净空区范围广、环境复杂，无人机目标具有低、慢、小等特点，对现有探测手段造成了严峻的挑战。为监视机场净空区非合作无人机目标，本项目结合无线电侦测、雷达和光电等传感器，在提高单传感器探测性能的基础上，重点研究了基于多传感器的协同探测技术、综合识别技术以及目标监视网络构建方法，并在机场净空区对部署方案、协同探测和综合识别技术进行了验证。针对无线电侦测，搭建了分布式TDOA系统，充分利用无人机信号的多域特征，可有效探测和识别复杂电磁环境中的消费级无人机机型超过一百种。针对雷达目标探测与识别，研究了基于回波波形信息的虚警抑制技术，有效降低了杂波引起的虚警；研究了综合多域特征的融合识别网络，实测数据验证雷达对无人机与飞鸟的识别准确率达到96.41%。针对光电目标探测与识别，研究了基于多帧的红外序列小目标检测算法，实现了低空目标的高效检测；研究了多尺度融合超分辨率重建技术，有效增强图像目标识别特征，提升识别距离达15%。在满足机场净空区严格电磁辐射要求的前提下，构建了以被动探测手段为主的多传感器协同探测与识别系统，以较低成本实现了对低慢小目标的高效探测与可靠识别。研究过程中建立了无线电侦测、雷达、光电图像目标识别数据集，在机场、监狱以及城市等重要场所形成了应用示范。研究成果为机场净空区低空安全监视系统的研制和应用奠定了技术基础，将提升机场对非合作无人机目标的探测和识别能力。