考虑通信延迟的多自治水下航行器协同定位算法

Cooperative localization algorithm considering of communication delay for autonomous underwater vehicles

中文关键词  自治水下航行器  协同定位  通信延迟  扩展卡尔曼滤波  误差  计算机仿真

英文关键词  autonomous underwater vehicles  cooperative localization  communication delay  extended Kalman filtering  errors  computer simulation

中文摘要

      自治水下航行器(AUV)协同定位中通信延迟具有常态性. 面对延迟到达的信息, 传统方法一般会有定位精 度或实时性的损失. 针对通信延迟的不利影响, 本文在建立水声探测和通信时延模型的基础上, 以扩展卡尔曼滤波 (EKF)为算法框架, 提出了信息顺序到达和信息出序到达2种协同定位算法, 并以建构面向信息出序情景的算法为 主要创新工作. 在信息顺序到达算法中, 将延迟信息进行序贯处理以减小定位误差. 在信息出序到达算法中, 以信 息出现一步滞后的延迟为背景, 使用出序信息直接对从AUV最新状态估计进行再更新, 信息无损地实时估计运动 状态. 计算机仿真实验结果表明, 本文算法相比于传统的航位推算、整周期滤波、量测丢弃等方法, 具有更高的估计 精度; 相比于数据缓存滤波、重新滤波等方法, 具有强实时性.

英文摘要

      Communication delay for the cooperative localization of autonomous underwater vehicles (AUV) occurs frequently. For information arrival delay, traditional methods generally have a loss in localization accuracy or real-time performance. Aiming at overcoming the adverse influence of communication delay, two cooperative localization algorithms, based on the establishment of underwater acoustic detection and communication delay models and under the extended Kalman filtering (EKF) algorithm framework, are proposed in this paper to cope with the probable situations: information sequential arrival and information out-of-sequence arrival. And the main innovation work is to construct the algorithm of information out-of-sequence. In the algorithm of information sequential arrival, the information is processed sequentially to reduce localization errors. In the case of information out-of-sequence arrival, the out-of-sequence information is used directly to further update the latest motion state of the slaver AUV, and the motion state is estimated without loss of information in real time under the background of one-step delay of the information. The computer simulation experimental results show that the proposed algorithms have higher estimation accuracy than the traditional methods such as the dead reckoning, the full-period filtering and the measurement discarding, and have strong real-time performance compared with the methods of the data cache filtering and re-filtering.