摘要
在基于波达方向估计(DOA)的多传感器节点目标三维空间定位算法中,当目标与传感器节点处于不同深度时,定位原理决定了不同的俯仰角误差将导致不同的定位误差,特别是当传感器与目标接近于同一平面,即俯仰角接近90°时,较小的俯仰角误差将引起非常大的定位误差。为了解决该问题,论文提出在最小二乘定位算法中,根据不同俯仰角进行加权的方法,降低俯仰角估计值接近90°的节点分量在定位方程中的贡献,减少俯仰角估计误差对综合定位精度的影响,提高定位算法对俯仰角估计误差的稳健性。文中对单个目标的定位进行了计算机仿真,对比了某些节点深度与目标接近情况下加权最小二乘算法和最小二乘算法的定位性能,仿真结果说明当参与定位的部分节点与目标深度接近时,所提加权最小二乘算法对目标位置的估计均方根误差低于最小二乘算法,具有更高的定位精度和稳健性。
In the three-dimensional localization algorithm based on direction-of-arrival(DOA),the target’s location errors are sensitive to the estimation errors of pitching angles,especially when the sensor array nodes are almost in the same horizontal plane with the target.In other words,the location errors are highly sensitive to the pitching angle errors when the pitching angles are close to 90°.To solve the sensitivity problem,a weighted least squares method is proposed in this paper.Small weighting factor is chosen for those sensor array nodes with large pitching angle estimation to decrease the impact of pitching angle estimation error and to increase robustness of the method.Furthermore,the performance of the weighted least squares method is analyzed via numerical simulation.Simulation results show that the present weighted least squares method has better robustness over DOA estimation errors and disturbance of the sensor array nodes than ordinary least squares method.
作者
杜金香
许恒博
祝鹏
DU Jin-xiang;XU Heng-bo;ZHU Peng(School of Marine Science and Technology,Northwestern Polytechnical University,Xi’an 710072,China)
出处
《水下无人系统学报》
北大核心
2019年第5期570-573,共4页
Journal of Unmanned Undersea Systems
基金
国家自然基金项目资助(61301197)
西北工业大学中央高校基本科研业务费基础研究基金(1069920140011)
水下信息与控制重点实验室开放基金(614221801040217)
关键词
传感器
目标定位
最小二乘
加权最小二乘
波达方向估计
sensor
target localization
least square
weighted least square
direction-of-arrival(DOA)
