论文部分内容阅读
在惯组系统级标定过程中,由于系统模型尤其是噪声统计特性的不确定性,常常造成较大的估计误差,严重时甚至导致滤波器发散。针对此问题,采用Two-Stage滤波思想,研究随机噪声干扰下系统不确定性偏差的最优滤波器设计(OTSKF),并在此基础上提出一种基于最优TSKF算法的快速次优滤波算法。理论分析表明该算法具有较小的运算量、良好的收敛性及抗扰动性。最后,将该算法应用于惯组系统级标定,通过一组自动化标定方案,实现了惯组的在线标定,实验结果验证了该算法的有效性。
In the process of system-level calibration of inertial system, the uncertainty of the system model, especially the statistical characteristics of the noise, often leads to large estimation errors and even filter dispersions in severe cases. In order to solve this problem, the optimal filter design (OTSKF) for system uncertainties under random noise interference is studied by using Two-Stage filtering theory. Based on this, a fast sub-optimal filtering algorithm based on optimal TSKF algorithm . Theoretical analysis shows that this algorithm has a small amount of computation, good convergence and anti-disturbance. Finally, the algorithm is applied to the system-level calibration of the inertial system. The calibration of the inertia system is realized through a set of automated calibration schemes. The experimental results verify the effectiveness of the algorithm.