Return to Session A6 Next Abstract

Session A6: Adaptive KF Techniques, Data Integrity, and Error Modeling

A Method of Inertial Integrated Navigation Based on Low Cost MEMS Sensors
WANG Zheng-chun, XIONG Zhi, LV Pin, XU Jian-xin, HUANG Xin, XU Li-min, Navigation Research Center, Nanjing University of Aeronautics & Astronautics, China
Location: Big Sur

Abstract: As early as 80 ~ 90 years in the 20th century, the concept of pedestrian navigation system(PNS) was put forward, which is an important area of navigation in the emerging branch. PNS is a kind of system for position and navigation that determines the current location of pedestrians. Based on pedestrian movement, it can record the walking trajectory and guide pedestrians to reach the destination. Because PNS has good adaptability, convenience and more accurate navigation information, it has wide application prospects in military and civil areas such as man combat, rescue, personnel monitoring, traffic travel and so on. PNS is a method of track reckoning based on strapdown inertial navigation system. However, inertial navigation system is prone to error accumulation, resulting in low positioning accuracy and practicability. With the rapid development of computer technology and communication technology, PNS began to enter the real practical stage. Low cost inertial measurement unit(IMU) has wide market and application value. Since low cost IMU device has low accuracy, and the inertial navigation system has accumulated error, and the system is easy to diverge, and the GPS signal can not be used in some cases, it is necessary to design a kind of navigation method to adapt low cost IMU.
In this paper, it is presented that a foot inertial navigation system which is based on low cost MEMS inertial sensors is presented. MEMS device has the advantages of small size, easy installation, low price and low power consumption, which is one of the ideal devices for pedestrian navigation. According to the characteristics of pedestrian zero-velocity during walking steady state and running non-stationary state, an inertial navigation method based on Zero Velocity Update (ZUPT) is adopted. The method uses the three-axis magnetic information to calculate the initial motion heading angle, determine whether it is in the zero-velocity state according to the accelerometer and gyro output, calculate the height change by the barometer output, and solve the velocity, position, and attitude based on the traditional strapdown algorithm. Since the heading angle is not observable at the time of zero-velocity correction, and there is an accumulation error in the course angle over time, this paper studies a kind of self observation algorithm for heading error based on zero-velocity. In the zero-velocity state, the heading angle in the latter moment is theoretically equal to the heading angle at the previous moment. The Kalman filter method is used to fuse the information and correct the heading in real time. Finally, the velocity and position computed by traditional strapdown algorithm, GPS position and heading observation angle of the foot touching the ground at zero-velocity are filtered by Kalman filter as the observation to correct the velocity, position and attitude angle error, so as to improve the navigation accuracy. Based on the implementation scheme of pedestrian navigation system proposed in this paper, the steady state of walking and non-stationary state of running were experimentally verified through the actual test data. The experimental results show that by using the integrated navigation method designed in this paper, the navigation accuracy can reach 95% by walking 240 meters in 3 minutes, for running, and the navigation accuracy can reach 95% by running 125 meters in 1 minute.The experimental results show that the proposed method can provide efficient, reliable and continuous position and navigation service.
Key words: pedestrian navigation; zero-velocity correction; Kalman Filter



Return to Session A6 Next Abstract