Research On Motion Sensing Via Magnet Array Scanning

In recent years,with the rapid development of the Internet of Things,the intelligent sensing technology has been widely used and the magnetic sensing plays an important role in it.The traditional work based on magnetic sensing usually uses the geomagnetic information of different positions for navigation and location combined with the inertial data,or uses the magnetic characteristics of a single magnet for interaction with other smart equipments and activity recognition.Based on the bipolarity of magnets,this paper innovatively uses magnet array to customize the magnetic field and perform research on the Motion Sensing mechanism via magnet array scanning.This paper mainly focus on the following two aspects of research:First,this paper study and propose the indoor pedestrian tracking mechanism based on magnet array scanning.Each magnet array is constructed with a unique binary code which lables the indoor connection of different spaces.When pedestrians cross these connections,the intelligent sensing devices embeded in their shoes scan the corresponding magnet arrays.By analysing the collected data,each binary code corresponding to the magnet arry can be recognized.In order to deal with the diversity of gait trajectory,a generalized analytical model based on the spce axis is proposed to analyse the magnetic induction intensity signal.By continuously identifying the magnet array code of pedestrians and combining with indoor map information,the indoor tracking can be effectively realized.The real experiment results show that the prototype system can achieve an average accuracy of 88.9% in identifying the magnet barcodes and an average accuracy of 93.1% for indoor pedestrian tracking.Second,this paper study and propose the equipment training action measurement mechanism based on magnet array scanning.The mechanism uses the equidistant mag-net array to calibrate the training action with several sections,which provides a reference for action measurement.During the equipment training,the user scans the magnet array near the equipment with an intelligent sensing device at the wrist.The device collects the corresponding magnetic data and inertial data.By analysing the training action,the action direction extraction model is proposed to decompose and analyse the magnetic induction intensity signal combinated with the inertial data.Thus,the mechanism measures the training action in a fine-grained way,and perform up-to-standard detection on the training action.The real experiment results show that the prototype system can detect the calibration points with an average error less than 2cm.Meanwhile the system can achieve an average accuracy of 90.3% and 81.2% in range standard detection and speed deviation detection of a single training action,respectively.While the average accuracy of the whole training action up-to-standard detection is 90.2%.