一种带遗忘因子的自适应卡尔曼滤波器及其在移动机器人中的应用

摘要

Because it is difficult to establish the exact system model of wheeled mobile robot during the measurement process, its tilt angle is usual y imprecise. This paper uses an adaptive Kalman filter with a forgetting factor to change the level of confidence of the present state estimation from the filter and improve the influence of the inaccuracy of noise statistical model in the system on the meas-urement results. Simulation results show that, when the measurement model can not used to measure the tilt anlge accurately, the a-daptive Kalman filter is used to improve the accuracy of the tilt angle, because its effect is better than the traditional Kalman filter.%在移动机器人姿态估计过程中，由于难以建立准确的系统模型，导致参数量测不准确。采用一种自适应卡尔曼滤波器，通过加入遗忘因子的方式改变滤波器对状态估计的信任程度，改善了系统中噪声统计模型不准确对测量结果的影响。仿真结果表明，当经验知识不足导致建立姿态测量模型不准确时，自适应卡尔曼滤波器的滤波效果优于传统卡尔曼滤波器，提高了姿态信号的估计精度。