应急救援作战等任务中的定位需求存在时间的突发性、地点的不确定性,以及环境的复杂性等特点,对定位系统部署的灵活性、规模的可扩展性、环境的适应性都提出了高要求。本文围绕面向应急定位任务的区域定位系统关键技术展开研究,旨在解决应急任务场景中信息和资源受限为区域定位系统的部署与运行带来的多项挑战,提升区域定位系统在应急定位任务中的适应性。针对定位节点能力受限导致测距时钟误差抑制效率低的问题,本文提出一种基于非线性估计器的无线电测距技术。该技术采用分层结构,通过将部分节点设置为被动接收节点,提升了有限带宽下的通信与测量效率。对于被动接收节点的距离测量,使用非线性距离估计器提取出了时钟误差影响相同的中间量进行处理以实现时钟误差抑制,提升了有限测量信息的利用效率,避免了时钟同步的复杂运算,提升了节点间的测距效率。针对节点规模增大时,定位系统空间基准自主建立精度低、效率低的问题,本文提出一种基于并行坐标下降法的分布式基准自主建立技术。该技术通过分析坐标下降法的收敛性条件,结合节点网络的独立集划分,支持分组的并行坐标更新,有效提升了分布式定位的并行度,进而提升了基准自主建立的效率。同时,通过对锚点坐标先验信息的深度融合,实现了分布式的绝对坐标获取与定位精度优化。针对在复杂运动和测距干扰影响下节点实时定位精度低的问题,本文提出一种基于误差模型调节的位置跟踪与精化技术。该技术在引入节点动静状态信息的基础上,通过调节噪声模型,用位置分布变化描述节点的复杂运动状态,联合估计测距偏差并改变缺失测量的信息权重,而后使用扩展卡尔曼滤波融合历史测量信息完成位置更新。该技术有效提升了定位系统对复杂运动和测距干扰的鲁棒性,改善了便携式节点的实时定位精度,增强了对应急任务的适应性。针对实时地图与定位系统融合时匹配与定位精度低的问题,本文提出一种激光雷达辅助的坐标配准技术。该技术引入节点的布设先验来建立环境物体与各节点的空间联系,通过对实时点云地图的处理与建模,提取出环境对节点位置分布的额外空间约束,融合节点间距离约束联合估计各节点位置。该技术为区域定位系统引入了丰富的环境信息,实现了有效的信息融合,提升了定位精度以及定位系统与实时地图的匹配程度,进一步拓展了系统的实用价值。
The localization requirement in emergency rescue or combat missions is characterized by the suddenness of the timing, uncertainty of the location, and complexity of the environment, which put forward high demands on the deployment flexibility, scale scalability, and environmental adaptability of the positioning system. This dissertation studies the key technologies of local positioning systems for emergency localization tasks, in order to address the challenges caused by the limited information and resources in emergency scenarios for the deployment and operation of local positioning systems, and to enhance the adaptability of local positioning systems in emergency localization tasks.To address the issue of inefficient clock error suppression in ranging due to limited capabilities of positioning nodes, a radio ranging technology based on nonlinear estimators is proposed. The technology employs a hierarchical structure to improve the communication and measurement efficiency under limited bandwidth by setting partial nodes as passive receiving nodes. For the distance measurement of the passive receiving nodes, a nonlinear distance estimator is adopted to extract the intermediate variables with the same clock error impact for processing to achieve clock error suppression, which enhances the efficiency of utilizing limited measurement information, avoids the complex clock synchronization operation, and improves the efficiency of inter-node ranging.To solve the problem of low-accuracy and low-efficiency of spatial reference autonomous establishment of the local positioning system when the system capacity increases, a distributed reference autonomous establishment technology based on parallel coordinate descent method is proposed. The technology supports parallel coordinate update in groups by analyzing the convergence condition of coordinate descent method and combining with the independent set division of node network, which effectively improves the parallelism of distributed localization, and then enhances the efficiency of reference autonomous establishment. Meanwhile, the distributed absolute coordinate acquisition and positioning accuracy optimization are achieved through the deep fusion of a priori information on anchor coordinates.To address the problem of low real-time localization accuracy of nodes under the influence of complex motions and ranging disturbances, a error model adjustment-based position tracking and refinement technology is proposed. Based on introducing node dynamic and static state information, this technology describes the complex motion state of nodes by adjusting the noise model, jointly estimates the range bias and changes the information weight of missing measurements. Then the position update is completed by fusing the historical measurement information using extended Kalman filtering. This technology effectively enhances the robustness of the positioning system to complex motion and ranging disturbances, improves the real-time positioning accuracy of portable nodes, and strengthens the adaptability to emergency tasks.Aiming at the problem of low matching and localization accuracy when fusing the real-time map with the positioning system, a coordinate alignment technology assisted by the light detection and ranging is proposed. The technology introduces a priori of node placement to establish the spatial connection between environmental objects with each node, extracts additional spatial constraints from the environment on the node location distribution by processing and modeling the real-time point cloud map, and fuses the distance constraints between nodes to jointly estimate each node location. This technology introduces rich environmental information to the local positioning system, realizes effective information fusion, improves the localization accuracy and the degree of matching between the positioning system and the real-time map, and further expands the practical value of the system.