基于有向图切换的INS/GPS组合导航VSMM算法

A VSMM Algorithm Based on Digraph Switching for INS/GPS Integrated Navigation

INS/GPS组合导航系统量测噪声统计特性随着载体机动、外界扰动等影响发生变化,传统的卡尔曼滤波方法对于这种量测噪声变化系统的滤波精度不高,甚至会出现发散现象。针对上述问题,将机动目标跟踪中的变结构多模型滤波算法引入INS/GPS组合导航滤波中,提出了一种基于有向图切换的变结构多模型INS/GPS组合导航算法,并对此算法进行了仿真。仿真结果表明:该算法能够适应系统量测噪声变化,定位误差小于交互式多模型算法,并有效地降低了滤波过程中的计算量。

The characteristics of the observation noise in INS/GPS integrated navigation system changes with the vehicle maneuvering and disturbance of application environment. Conventional Kal- man filtering algorithm cannot get satisfactory accuracy, even divergent for this observation noise changing system. Aiming at this problem, this paper introduces the variable structure multiple model （VSMM）algorithm which is often used in maneuvering target tracking into filtering processing of the 1NS/GPS integrated navigation. Then a novel VSMM integrated filtering algorithm based on digraph switching （DS-VSMM） is presented. Simulation results show that DS-VSMM can adapt to the change of the system observation noise. DS-VSMM can get the same navigation accuracy as the interacting multiple model （IMM） algorithm with lower computation burden.