基于随机集理论的分布式异类传感器网络目标跟踪技术

This project mainly deals with random finite set (RFS) theory based target tracking over distributed heterogeneous sensor network (DH-SN). The methodology of target tracking over DH-SN can be established and enhanced by studying the following fundamental problems, including the modelling of conjugate prior for generic observation model (GOM), the modelling of information entropy balanced local posterior distribution, GOM based filtering method of single sensor node and posterior information fusion of multiple nodes towards the problem of inaccuracy information entropy. We will solve the key technologies, and propose an RFS theory based target tracking algorithm for DH-SN. Besides, simulation experiments and also real-world data will be used to validate the presented theory results and technological achievements. The research of this project is of great significance, since it can provide the theoretical and technical basis for the applications of target tracking technique in the real-world multi-sensor surveillance system. Moreover, the theoretical results can also be applied to target surveillance in the fields such as air defense network, autonomous vehicle technology and Internet of things.

本项目旨在解决基于随机集理论的分布式异类传感器网络（DH-SN）目标跟踪问题，通过研究广义量测模型下的共轭先验分布建模、信息熵均衡的本地后验分布建模、基于广义量测模型的单节点滤波方法、针对信息熵失准问题的多节点信息融合方法等基本问题，建立和完善DH-SN目标跟踪技术的理论体系；突破关键技术，建立一套基于随机集理论的DH-SN目标跟踪信号处理算法，设计仿真实验并利用实测数据，对本项目所获得的理论和技术成果进行验证。该项目研究可为目标跟踪技术推广应用到实际多传感器监视系统中奠定理论和技术基础，具有重要的科学意义；其理论和技术成果可以应用于国土防空、无人驾驶技术和物联网等领域的目标监视问题。

项目执行期间，按照研究目标，开展了展基于广义量测模型的单节点滤波和针对熵失准的多节点后验信息融合等研究内容，完成了基于多假设-多维联合共轭先验的多目标跟踪算法、基于多目标无信息熵分布的全局视域目标信息融合模型两个关键问题突破，最终形成了一套性能优异且稳健的分布式异类传感器网络目标跟踪技术，可扩大分布式异类传感器网络的探测视域，实现全局视域内多目标航迹精准稳健估计。最终，通过实验验证了关键技术成果，相关技术成果取得了报告、专利和学术论文等一系列技术成果。总体而言，完成了项目预期目标，为目标跟踪技术应用到实际传感器网络中奠定理论和技术基础。

内容获取失败，请点击重试