基于扩展卡尔曼滤波的高程估计算法

Height Estimation Algorithm Based on Extended Kalman Filter

在室内行人定位系统中,行人的高程定位精度关系到整个定位系统的可靠性。提出一种基于腰间传感器的室内行人高程估计算法。首先利用支持向量机识别行人上楼下楼动作,针对行人的运动状态采用自适应的高程估计算法。针对气压计测量值易受环境影响的问题,采用了基于EKF融合气压和加速度的高度估计算法,提高了高度估计算法的稳定性。经实验验证,当室内人员进行平地走、上楼等一连串动作后,基于差分气压测高法计算的高度误差为9.92%,基于加速度估计的行人高度误差为9.52%,EKF融合后定位误差下降到2.32%,提高了高程估计的精度。

In the indoor pedestrian positioning system,accuracy of height estimation is related to the reliability of the positioning system. A height estimation algorithm was proposed to accurately estimate pedestrian＇s height in a multi-floor building using a waist-mount device. The proposed algorithm extracts the features and adopts an effective stairway detection algorithm at first. Then EKF is used to calculate the vertical displacement of the pedestrian by fusing inertial measurements and barometric pressure measurements. The experiments show that the proposed algorithm is stable and error of the algorithm decreases to 2. 32% while the error of pressure-based method is 9. 92%and the error of accelerometer-based method is 9. 52%.