一种NLOS量测平滑算法在MAUVs协同定位中的应用
A NLOS Measurement Smoothing Algorithm for Cooperative Localization in Multiple Autonomous Underwater Vehicles

在基于双领航者的MAUVs协同定位过程中,为了减轻NLOS量测误差的影响,在假设NLOS修正偏差先验已知的前提下,以4状态Markov链描述了4种LOS/NLOS量测模型间相互独立转换过程,继而利用交互多模和Kalman滤波理论设计了一种AUVs间相对距离量测平滑算法,并将其距离量测估计结果应用于MAUVs协同定位系统中。仿真结果对比表明,该算法可以有效提高AUVs间的相对距离量测估计精度,获得了更好的协同定位性能。
In the process of MAUVs cooperative localization based on two leaders, in order to mitigate the influence of NLOS measurement deviation, a four state Markov chain is used to describe the switch process among four LOS/NLOS range measurement models which are independent of each other. Then we design a relative range measurement smoothing algorithm by combining the IMM and Kalman filter theory among AUVs, and the range measurement estimation results are applied to the MAUVs cooperative localization system. Simulation results and their analysis indicate preliminarily that the proposed algorithm does improve effectively the estimation accuracy of relative range measurement among AUVs. Moreover, the performance of the corresponding cooperative localization is also better than that of conventional method