基于改进的联邦UKF无人艇组合导航系统设计

Design of Integrated Navigation System for Unmanned Surface V ehicle Based on Improved Federated UKF

针对无人艇在高海况下长航时,大幅度作业滤波精度较低的问题,提出一种基于联邦结构的无迹卡尔曼滤波(Unscented Kalman Filtering,UKF)算法,将其应用于自主研制的无人艇组合导航系统中。建立系统误差方程和量测方程;引入渐消因子、基于量测值与预测量测值差值的可变因子和自适应最优信息分配因子对联邦UKF算法进行改进,保持信息的强跟踪特性和组合导航系统的信息融合精度,得到全局最优估计值。开展湖试试验,验证该组合导航系统的有效性,结果表明该系统实时性、稳定性好,抗干扰能力强,能有效提高导航精度。该方法不仅能为无人艇作业提供安全保障,而且可供其他组合导航系统设计参考。

An Unscented Kalman Filtering(UKF)algorithm based on the improved federated structure is proposed to address the problem of low filtering accuracy in largescale operations of unmanned boats in the long voyages under high sea conditions,which is applied to a selfdeveloped unmanned boat integrated navigation system.Systematic error equations and measurement equations are established.The federated UKF algorithm is improved by introducing the fading factor,the variable factor based on the difference between measured and predicted values,and the adaptive optimal information allocation factor to maintain strong information tracking and information fusion accuracy of the integrated navigation system,and to obtain the global optimal estimate.A lake trial was conducted to verify the effectiveness of the integrated navigation system,and the results showed that the system has good realtime performance,stability,and strong antiinterference ability,which effectively improves navigation accuracy.This method not only provides safety assurance for unmanned craft operation,but also serves as a reference for the design of other integrated navigation systems.