Position Tracking Technology Research Based On Strapdown Inertial Navigation

More and more applications need to track on people, especially the people who works outdoor and indoor. in the field of medical virtual reality and game, access to data in a human body movement, can help to describe the human body movement and obtain more information on potential. In clinical rehabilitation, Tracking the movements of patients in outdoor environment can real-time monitor the activities of the patients and correct wrong actions in timely; in the field of game, create a realistic virtual world and make the game objects and characters have realistic movement is a key to the success of the animation; in the field of interpersonal interaction, intelligent machines not only need communicate with people by voice directly, but also need to have posture interaction which is more naturally. In more areas, people expect learn more about how much sport time spent in daily life, running information such as the path to further adjust sport regularity in their activities.In most cases, there are much of repeated and identifiable cycle within the human body movement pattern, among the movement of the acceleration and angular velocity measurements is zero value at a certain time point. these zero value appears before the swing phase of the gait cycle, in general, this period is called stand phase. In this period, the feet of human body contact with the ground. Identify the periodic law of gait can help us made less error in tracking signal, estimate the more accurate location information. In order to make the precision of positioning is not easily affected by the outside world and do not depend on other infrastructure support. A kind of inertial sensor installed on the human body is good for position tracking in the movement. With this kind of inertial sensor based system has the advantages of fully independent positioning. Measurement unit is not influenced by motion model, also won't be affected by the location environment, thus has received the widespread attention. However, low cost of gyroscope devices will be affected by temperature,so the error in the output signal will accumulate with time.This makes the positioning information is not accurate, so it is badly needed a plan for low computation and high positioning accuracy.This paper begins with a brief introduction and analysis of attitude estimation technology based on strapdown inertial navigation, elaborates the base principle of quaternion algorithm, show the advantages of attitude algorithm that acceleration and magnetometers to compensate the error by contrast with the performance of traditional quaternion algorithm in recording and tracking of moving objects. On this basis, I have been deeply studied on time complexity and performance of different attitude algorithm, the attitude algorithm suitable for the pacers algorithm is proposed. Next, considering that the characteristics of the pacers motion sensor signal is periodic change. studied the the gait change rule. The gait cycle calculation based on a frequency domain and zero velocity update algorithm is given. Finally, in a real environment, using orthogonal accelerometer, orthogonal gyroscope and three axis magnetometer collect real-time data, use difference method to extract the effective data fragment and calculate data through multiple sampling rate.Research results show that after pretreatment, within the different time slot on the data, using different gesture decoding algorithm, can effectively reduce the computation complexity and more smooth tracking trajectory. coupled with the characteristics of the movement patterns, install the sensor on the shoes to track pacers trajectory, by the method of zero velocity update to further improve tracking precision.