具有协方差反馈控制的无人机INS/GNSS组合自适应CKF算法

An Adaptive CKF with Covariance Feedback Control for UAV INS/GNSS Integration

为提高无人机INS/GNSS组合导航对过程噪声方差的自适应能力,提出了一种具有先验状态协方差反馈控制的自适应容积卡尔曼滤波(CKF)算法。在该算法中,首先将系统后验状态和误差协方差信息反馈至滤波过程中,构成CKF协方差传播的闭环结构;然后,基于极大似然准则,利用估计窗口内的反馈状态和误差协方差信息,建立了一种先验状态协方差在线反馈控制策略。提出的方法在调整过程中能够保证先验状态协方差的正定性,克服了传统噪声统计估计方法需要对负定结果进行额外修正的缺陷,从而有效抑制了过程噪声方差不确定性对滤波解的影响,提高了CKF用于无人机INS/GNSS组合导航解算时的自适应能力。通过无人机INS/GNSS组合导航仿真实验验证了提出算法的有效性。

This paper presents an adaptive Cubature Kalman Filter(CKF)with the feedback control of prior state covariance to improve the adaptive ability of INS/GNSS integration against the uncertainty involved in process noise covariance.Firstly,the system’s posterior state and its error covariance are fed back to the filtering process to form a closed-loop structure for CKF covariance propagation;then,a prior state covariance feedback control strategy is established based on the maximum likelihood principle using the feedback state and error covariance information within an estimation window.The proposed method can ensure the positive definiteness of prior state covariance in the adjustment process,and at the same time it can overcome the defect of the traditional noise statistical estimation methods that they need additional correction to deal with the negative definite result.Thus,the proposed methodology can effectively suppress the influence of process noise covariance uncertainty on the filtering solution,and improve the adaptive ability of CKF when applied in the solution of INS/GNSS integration.Its effectiveness is proved through the simulation of INS/GNSS integration for UAV navigation.