Research on Pedestrian Location Based on Dual MIMU/ Magnetometer/ Ultrasonic Module
Wang Giuying, Guo Zheng, Zhang Minghui, Cui Xufei, Harbin Engineering University, China; Wu Hui, Dalian Shipbuilding Industry Offshore Co., Ltd, China;Jia Li, Department of Electrical and Computer Engineering, Oakland University, USA.
Location: Big Sur
Inertial positioning technology is one of the most important positioning techniques. Its advantages are low power consumption, no radiation, integrating easily, and requiring no external assistance information. It is suitable for the unknown environment without GPS, lack of radio base stations or have no a priori geographic information database. Inertial positioning technology plays an important role in disaster relief and man combat. The pedestrian location system based on the micro inertial measurement unit (MIMU) is usually integrated in the foot of the pedestrian, which is more effective than in the lumbar of the pedestrian, using the inertial law to solve the pedestrian's position, velocity and attitude information. What’s more, the inertial location system utilize the zero velocity update (ZUPTs) algorithm to suppress their errors. As a result of ZUPTs, the positioning accuracy is increased, especially the speed information, roll angle and pitch angle. However, the zero velocity update method cannot effectively decrease pedestrian position errors and heading angle error. So pedestrian’s position errors and heading angle error will still accumulate over time. The magnetometer can detect Earth's magnetic field information and solving the heading angle by it, but only computing the heading angle information cannot achieve positioning. And man-made magnetic field is not applicable to the unknown and disastrous environment. The range of ultrasonic positioning is smaller than others and also depends on the station construction. So ultrasonic positioning does not apply to environments where the field and prior information are unknown. In this paper, a dual MIMU/ magnetometer/ ultrasonic pedestrian location scheme is proposed. The zero velocity update algorithm and the Kalman filter algorithm based on the equality constraint are used to correct the position, velocity and attitude information of the pedestrian. The two MIMUs, two magnetometers and two ultrasonic modules are respectively fixed on the left and right foot of pedestrians, and the output information of magnetometer is used to assist the zero velocity update algorithm to calculate the position information and attitude information of the pedestrian, which improves the observability of pedestrian heading angle and suppresses heading error. We use two ultrasonic modules to measure the spatial relative position relation of pedestrian’s left foot and right foot in real time and construct the second order nonlinear equation constraint equation. We correct pedestrian location information continually by using the constraints so that to improve the positioning accuracy. To verify this program, in this paper, a practical pedestrian positioning system is designed. And we have designed different pedestrian trajectories in the building and the playground, such as straight lines, rectangles, and so on. After the experiment we processed and analyzed the offline data from the dual MIMU/ magnetometer/ ultrasonic module positioning system integrated on the foot. The experimental results show that compared with single MIMU pedestrian location scheme or ultrasonic location scheme, pedestrian positioning technology based on dual MIMU/ magnetometer/ ultrasonic module has higher positioning accuracy, and the positioning accuracy of this system is better in the longer time, which can meet the requirement of pedestrian positioning where there is no external auxiliary information.