Inertial Measurement Units Based Hand Position Tracking Technology Research

Cerebral apoplexy is also known as stroke.In recent years,the incidence of this disease has gradually increased,and most of the patients caught the disease of limb hemiplegia,of which upper limb hemiplegia accounted for a large part,to some extent,rehabilitation training can slow the patients’ disease.However,the effect of traditional rehabilitation training depends on the guidance of rehabilitation doctors too much.In recent years,assisted rehabilitation training methods have been greatly developed,including mechanical motion tracking system,optical motion tracking system,electromagnetic motion tracking system,acoustic motion tracking system and video motion tracking system.However,these systems have the disadvantages of high installation cost,limited training to the patient and high price,and the accuracy of the system needs to be improved.The fast development of micro-electro-mechanical systems(MEMS)with low power cost and small volume has greatly promoted the progress of inertial sensing technology.Inertial navigation tracking systems,which are based on Inertial Measurement Units(IMU)with MEMS,are used more frequently in the field of human rehabilitation recently.Inertial navigation system has the advantage of high positioning accuracy in short time,small size,light weight,less restrictions on the patient’s motion,low cost.Based on this,this paper plans to design an upper limb position tracking program based on inertial measurement unit,record the speed,position,posture and other information of upper limbs during rehabilitation training,guide and correct the patient training according to the requirements of standard rehabilitation training.Inertial navigation system has the advantage of high positioning accuracy in short time.But in long time,the positioning accuracy decreased drastically because of the cumulative error.In response to this feature,this article has done the following major work to improve system accuracy:1.The strapdown inertial navigation algorithm is designed,the quaternion algorithm is used to update the attitude,at the same time,the acceleration information is converted into the carrier coordinate system according to the coordinate system conversion relationship,and the information such as the target speed and position isintegrated.2.The zero speed correction technique is introduced according to the characteristics of the cyclical zero period of flexion and extension movement in the standard rehabilitation training arm.The stance hypothesis optimal detector is used to detect the moment when the target velocity is zero and trigger Kalman filtering.3.The "zero speed" information of the zero-speed interval is used as the measurement value for Kalman filtering.Then,the estimated results are fed back to correct the cumulative error of the strapdown inertial navigation.4.At the same time,a fixed interval smoothing algorithm(Rauch-Tung-Striebel,RTS)is designed on the basis of Kalman filter,which solves the problem of abrupt change of motion trajectory caused by zero speed correction.The navigation scheme based on the Kalman filter can effectively track the upper limb position of the patient’s rehabilitation training through simulations and experiments.For instance,in 108 seconds,the positioning error is 0.089% of the moving distance,meeting the accuracy requirement of auxiliary rehabilitation training.