基于鞋载惯性传感器的地图辅助人员定位方法

由于严重的“遮蔽”现象,卫星导航在室内环境中几乎无法使用。人员自主定位系统作为此条件下的一种有效“补盲”手段,在消防、特战、反恐等方面具有广泛的应用。为了克服人员自主定位中存在的惯性漂移,常常利用辅助信息提高定位精度。在粒子滤波算法框架下,提出一种改进的地图辅助人员定位方法。该方法通过引入粒子的方向加权,充分利用地图信息;克服了传统方法中由地图信息错误导致定位失败的缺点,提高了定位算法的稳健性。通过传感器的实测数据对算法进行验证,新方法随时间变化可保持1 m的定位精度,相对于传统方法有较大提高。

Map-aided pedestrian navigation based on foot-mounted inertial sensors

The satellite navigation signal is barely available due to the obscuration indoors. In that condition, the inertial-based system is an effective replacement for pedestrian navigation and has a wide range of applications, such as fire-fighting, special-operation and anti-terrorism. To suppress the drift in inertial based pedestrian navigation, many different types of aiding knowledge are adopted. An improved map-aided pedestrian navigation method is proposed under a particle filter framework. In the method, particles are weighted according to the direction information incorporated in the maps, and the location robustness is enhanced by recovering from location failure when partially wrong map information appeared. Based on the author-developed system, several experiments are conducted. The results manifest that the new method can maintain an accuracy of 1 meter as time increases, which is a great improvement on the traditional method.