基于MEMS惯性器件的行人室内定位系统

An Indoor Pedestrian Positioning System Based on MEMS Inertial Devices

摘要:

基于惯性传感器的行人航位推算系统不需要预先安装任何基础设备,能自主运行、实现实时行人定位。设计的硬件平台将低成本、低功耗、小尺寸的MEMS惯性传感器与GPS接收机相结合。在室内、城市峡谷等GPS信号不稳定的环境,惯性传感器根据前一GPS定点推算行人行走的相对位置。行人所处位置高度由气压计测量,与平面位置相结合实现三维定位。简单而有效的跨步探测及步长估计算法降低对微处理器的计算及存储要求。利用互补滤波器融合加速度计、陀螺仪、数字罗盘数据,降低方位误差、提高定位精度。室内行人行走测试实验表明:定位误差低于总行走距离的3%。验证了系统的准确性和可靠性,满足行人定位要求。

Abstract:

The pedestrian dead reckoning system based on the inertial sensors can autonomously operate and achieve the real-time pedestrian positioning, which does not need any pre-installed infrastructures. The designed hardware platform integrates the low-cost, low-power, small-size MEMS inertial sensors and GPS receiver. When the GPS signal is unavailable, such as indoor and urban canyons environments, the inertial sensors calculate the relative position of pedestrians from the previous GPS fixed point. The height information is measured by the barometer, which is combined with the plane position to realize the three-dimensional positioning. A simple but effective algorithm for stride detection and step length estimation is designed to reduce the calculation and storage requirements to the microprocessor. The proposed complementary filter fuses the accelerometer, gyroscope and digital compass data to reduce azimuth error and improve positioning accuracy. The indoor pedestrian walking test experiments show that the positioning error is less than 3% of the total traveled distance. The accuracy and reliability of the system is verified, which meet the requirements of the pedestrian position.