On a novel combination method for pedestrian multi-attitude solution with the wearable MEMS-IMU

Pedestrian's attitude solution plays an important role in a lot ofapplication areas. For the problem ofthe pedestrian's wearable attitude solution, two kinds of wearable methods, chest-wom and waist-worn, are proposed with a novel combination method including Zero momentary attitude detection + Geomagnetic correction + Third-order Taylor expansion quaternion update algorithm (ZGT) to resolve the pedestrian's attitude. In addition, a walking experiment is designed on the multi-attitude such as snake-like, pitching, wobbling, spinning and moving backward. MEMS-IMU and laptop are used for the test data output, collecting, recording and processing the data. After compared with the other methods, the IMU placed on the waist and with new combination ZGT is selected as the best solution for this research. This concise method can achieve easily the 14000 continuous data processed in 5.01s and the total attitude angle deviation of the mean magnitude up to 10
^-3
° compared with the result of international advanced production sbg MEMS-IMU as an approximate true value standard by itself own complicated and precise algorithm. In other words, it also suggested that feasibility and effectiveness of this method. Applied to the outdoor personnel positioning area, from calculating the difference between start and end point by moving an integral loop path on the track figure we can see: horizontal error accuracy, foot-mounted IMU with this method, can reach 0.8% in a total 230m travel. ZGT also performs well on heading correction with waist-wom IMU reducing the 9.1° yaw deviation in another experiment. To the further applications, optimization and derivation, it can be considered as a new combined solution in the other fields when use the free inertial navigation which has accumulated error solely.