超高速目标协同探测网络的非正常数据检测与柔性重构

Ultra-high-speed aircrafts like near-space vehicles have posed a great challenge to existing detection and early warning systems. This project takes the cooperative detection of ultra-high-speed targets with high mobility and intense antagonism as the background, considering the real-time performance, heterogeneity, multi-source, and reliability of networked collaborative detection data, etc. In the project we will build a distributed parallel collaborative flexible filter framework taking into account the data fusion process, tracking filter, non-normal data analysis, fault diagnose and flexible reorganization. We else adopt the strategy of real-time online collaborative fault diagnosis to detect, diagnosis, and isolate the deceptive data and interference data produced by the target, failure data and redundant data produced by detecting nodes, communication network failure data and other non-normal data. Flexible restruction methods are utilized to isolate the fault nodes, eliminate fraud and malicious data, reconstruct non-normal data and reduce the influence of non-normal data on the whole detection system. These methods can increase the collaborative detection probability of targets with ultra-high speed using three-dimensional sensor network in complex confrontation conditions, improve the quickness, accuracy and reliability of detection systems, and lay an important foundation for detecting ultra-high-speed targets in the future.

以临近空间飞行器为代表的超高速飞行器已对现有的探测预警体系构成了严峻挑战。本项目针对具有高机动性、强对抗性特点的超高速目标的协同探测问题，考虑网络化协同探测数据的实时性、异构性、多源性和可靠性等问题，建立兼顾数据融合处理、跟踪滤波、非正常数据分析、故障诊断和柔性重组的分布式并行协同柔性滤波器框架；采用在线协同故障诊断策略检测、诊断和隔离探测数据中由目标产生的欺骗数据和干扰数据，探测节点产生的故障数据和冗余数据，以及由于通信网络限制出现故障数据等非正常数据；进一步利用柔性重构方法隔离故障节点，剔除欺骗数据，重构非正常数据，降低非正常数据对整个探测体系的影响。提高在复杂对抗条件下三维传感器网络对超高速目标的协同探测概率及探测体系的快速性、精确性和可靠性，为我国未来在面向超高速目标的探测对抗中提供基础研究支持。

以临近空间飞行器为代表的超高速飞行器已对现有的探测预警体系构成了严峻挑战。本项目针对具有高机动性、强对抗性特点的超高速目标的协同探测问题，考虑网络化协同探测数据的实时性、异构性、多源性和可靠性等问题，建立兼顾数据融合处理、跟踪滤波、非正常数据分析、故障诊断和柔性重组的分布式并行协同柔性滤波器框架；采用在线协同故障诊断策略检测、诊断和隔离探测数据中由目标产生的欺骗数据和干扰数据，探测节点产生的故障数据和冗余数据，以及由于通信网络限制出现故障数据等非正常数据；进一步利用柔性重构方法隔离故障节点，剔除欺骗数据，重构非正常数据，降低非正常数据对整个探测体系的影响。提高在复杂对抗条件下三维传感器网络对超高速目标的协同探测概率及探测体系的快速性、精确性和可靠性，为我国未来在面向超高速目标的探测对抗中提供基础研究支持。项目批准以来，项目组共发表标注项目支持的SCI期刊论文4篇，EI论文10篇。申请发明专利2项，授权发明专利16项，撰写国防报告1篇。获得国防科技进步二等奖2项，新申请相关项目2项。