An Autonomous Waist-Mounted Pedestrian Dead Reckoning System by Coupling Low-Cost MEMS Inertial Sensors and GPS Receiver for 3D Urban Navigation

Abstract

Global positioning system (GPS) offers a perfect solution to the 3-dimension(3D) navigation. However, the GPS-only solution can’t provide continuous and accurate position information in the unfavourable environments, such as urban canyons, indoor buildings, dense foliages due to signal blockage, interference, or jamming etc. A pedestrian dead reckoning (PDR) system integrating the self-contained inertial sensors with GPS receiver is proposed to provide a seamless outdoor/indoor 3D pedestrian navigation. The MEM sensor module attached to the user’s waist is composed of a 3-axis accelerometer, a 3-axis gyroscope, a 3-axis digital compass and a barometric pressure sensor, which doesn’t rely on any infrastructure. The positioning algorithm implements a loosely coupled GPS/PDR integration. The sensor data are fused via a complementary filter to reduce the integral drift and magnetic disturbance for accurate heading. The four key components of the PDR algorithm: step detection, stride length estimation, heading and position determination are described in detail and implemented by the microcontroller. The step is detected using the accelerometer signals by the combination of three approaches: sliding window, peak detection and zero-crossing. The step length is estimated using a simple linear relationship with the step frequency. By coupling the step length, azimuth and height, 3D navigation is achieved. The performance of the proposed system is carefully verified through several field outdoor and indoor walking tests. The positioning errors are below 3% of the total traveled distance. The main error source comes from the orientation estimation. The results indicate that the proposed system is effective in accurate tracking.

Keywords

• pedestrian dead reckoning (PDR)
• indoor positioning
• inertial navigation system
• complementary filter
• sensor fusion