复杂空间约束下目标运动建模及跟踪方法研究

There exist a lot of spatial constrained problems in practical applications, for example, target’s motions are strictly constrained by the roads on the ground, the channels in the rivers, ship routes on the sea, air lines and so on. These spatial constraints contain prior information about target’s position and velocity. However, the conventional tracking theory uses unconstrained motion models, which are unable in describing complex motions in complex constraints accurately, and consequently fails to make full use of the significant prior knowledge contained in the spatial constraints. This leads to serious information waste and performance degradation. This project focuses on the problem of target tracking with complex spatial constraints. Under the framework of system modeling, where the knowledge of spatial constraint and target motion are considered independently, the 1-dimensional dynamic model sets which exactly cover the motion characteristics of target within spatial constraints will be studied, as well as accurate and effective static modeling of constraints. These works will solve the difficult problem of complex motion modeling with complex constraints. Based on the resulted models, the constraint coordinate state estimation theory with high nonlinearity, non-Gaussian and/or varying models, and the inner mechanism of performance improvement rising from incorporation of constraints are to be explored. Finally, data association with multiple uncertainties, track management and other problems concerned with multiple target tracking are also to be discussed, and a performance evaluation system for tracking with spatial constraints is also to be established. The main purpose of this proposal is to break through key techniques concerned with dynamic system modeling and tracking in constraint coordinates, and develop a fairly perfect tracking principle with spatial constraints. The resulting theories and techniques will provide strong support for the various real applications to approach optimal performance.

在实际跟踪应用中，存在大量空间约束问题，如目标运动被限制在地面道路、江河航道、海空航线等狭长区域。空间约束隐含了关于目标位置和速度的先验信息，然而传统跟踪理论使用常规无约束目标模型，无法给出复杂约束空间复杂运动的准确描述，未能利用空间约束这一重要的先验知识，造成了严重的信息浪费和显著的性能损失。本项目研究复杂空间约束下的目标跟踪问题，基于轨迹形状与运动规律分离建模的思想，研究准确高效的静态建模方法，探索反映约束空间运动特点的一维动态模型，破解复杂空间约束复杂运动建模的难题；研究强非线性、非高斯、模型复杂多变等新情况下的约束坐标状态估计理论，探索约束信息改善跟踪性能的内在机理；研究空间约束多重不确定因素下的数据关联、航迹管理等多目标跟踪问题，并构建空间约束跟踪性能评价体系；突破空间约束运动建模与跟踪的若干关键技术，提出较为完善的空间约束跟踪理论与方法，为大量应用系统逼近最优性能提供理论支撑。

在实际跟踪应用中，存在大量空间约束问题，如目标运动被限制在地面道路、江河航道、海空航线等狭长区域。空间约束隐含了关于目标位置和速度的先验信息，然而传统跟踪理论使用常规无约束目标模型，无法给出复杂约束空间复杂运动的准确描述，未能利用空间约束这一重要的先验知识，造成了严重的信息浪费和显著的性能损失。本项目研究复杂空间约束下的目标跟踪问题，基于轨迹形状与运动规律分离建模的思想，研究准确高效的静态建模方法，探索反映约束空间运动特点的一维动态模型，破解复杂空间约束复杂运动建模的难题；研究强非线性、非高斯、模型复杂多变等新情况下的约束坐标状态估计理论，探索约束信息改善跟踪性能的内在机理；研究空间约束多重不确定因素下的数据关联、航迹管理等多目标跟踪问题，并构建空间约束跟踪性能评价体系；突破空间约束运动建模与跟踪的若干关键技术，提出较为完善的空间约束跟踪理论与方法，为大量应用系统逼近最优性能提供理论支撑。经过4年的研究，本项目提出了目的地约束、航向约束、轨迹特征约束等新概念，建立了提取和利用隐含在非完整条件中的宝贵约束信息，以提升道路目标、反辐射导弹、航路目标等场景的目标跟踪性能；针对复杂空间约束问题，分别研究了直线轨迹和圆弧轨迹的隐式约束建模与跟踪问题，并提出了两种轨迹联接的约束建模与跟踪方法；此外，提出了基于静态-动态分离建模与跟踪、空时联合多目标跟踪等创新方法，建立了较为完善的空间约束跟踪理论与方法。在IEEE Transactions等JCR二区以上期刊发表学术论文15篇，其中JCR一区论文14篇；在国内外主流学术会议上发表学术论文24篇；申请并获授权发明专利20余项。获中国专利银奖1项，培养硕士生8人，博士生10人，指导学生获中国电子学会优秀硕士学位论文提名奖1项，多人获国家奖学金或行业特别奖学金。