【摘要】 当GPS卫星信号受到遮挡,车载GPS无定位输出时,通过低成本MEMS惯导进行定位是车辆导航的一种弥补方法。然而低成本MEMS惯导测量误差大,定位误差会快速累积。针对此问题,研究了一种应用于MEMS惯导导航定位的Kalman滤波算法。通过分析车辆的运动特性,在已有的研究基础上,提出了向心加速度差值误差这一新型观测量,并推导了误差状态系统模型的状态转移矩阵和观测矩阵。200 s时长的实车实验表明,单纯MEMS惯导定位的误差率是75.27%,而所述新方法的定位误差率是3.86%,定位精度有了大幅度提高,取得了良好的效果。更多还原

【Abstract】 When vehicle GPS has no positioning output due to GPS satellite signal’s being blocked, a positioning method based on low-cost MEMS IMU can be used to compensate the vehicle navigation. However, the measurement error of the low-cost MEMS IMU is large, and its positioning error would accumulate quickly. To solve this problem, a Kalman filtering algorithm for MEMS IMU navigation was studied. By analyzing the motion characteristics of vehicle, a new observation, i.e. the error of the centripetal acceleration difference, was presented. Then the state transition matrix and observation matrix of the error state system model were deduced. The real vehicle experiment for 200 s shows that the positioning error rate of autonomous MEMS IMU is 75.27%, and the positioning error rate by this new method is 3.86%, which show that the positioning accuracy has been significantly improved.更多还原