Research On Human Arm Motion Tracking System Based On MEMS Sensor

With the development of MEMS,computer and information technology,robot technology has been rapidly developed and widely used.Robot teaching is one of the main ways to achieve robot trajectory reproduction.The conventional teaching method has many problems,such as complicated teaching process,low efficiency,high learning cost and no security guarantee.To solve these problems,this paper presents a human arm motion tracking system and its method based on MEMS inertial sensor.An inertial sensor is used to obtain the relevant data of arm motion.Based on the theory of inertial navigation,an improved attitude solution method is proposed,and the effectiveness of the improved method is verified by simulation experiments.The position and trajectory of the end of the instructor’s arm in three-dimensional space were calculated by the method of attitude estimation position and spatial data superposition.The robot completed teaching and learning according to the motion trajectory of the arm.This method can reduce the teaching cost of the robot,improve the efficiency of teaching and ensure the safety of the instructor.The main work and innovation points are as follows:(1)This paper expounds the background and significance of applying the human arm motion tracking method based on MEMS sensor to robot teaching,introduces the development status of MEMS sensor used in human motion analysis,constructs the geometric model of the arm,and analyzes the advantages and disadvantages of Euler Angle method,quaternion method and direction cosine method in attitude solving.Quaternion method is chosen as the main gesture representation method in this paper.(2)This paper introduces the working principle of MEMS sensor,analyzes its noise characteristics,and constructs the median filter and mean filter algorithms to preprocess the original data.The simulation results show that the noise signal intensity is significantly reduced after filtering;The complementary filtering algorithm and extended kalman filtering algorithm are designed for two-dimensional palm motion,and the performance of the two algorithms is compared and analyzed by simulation experiments.The simulation results show that the attitude Angle accuracy of the latter is better.(3)To solve the problem of low accuracy of the attitude angle of the arm movement by the traditional attitude solving method,the adaptive volume kalman filter algorithm was introduced.In order to improve the accuracy and stability of the system in attitude solving,a maximum posterior noise estimator was introduced to estimate the system noise online,and an attitude solving method based on the combination of complementary filter and adaptive volume kalman filter was proposed.Under static and dynamic conditions,the improved method is compared with the traditional attitude solving method through simulation experiments.The results show that the proposed method has higher accuracy and better stability.(4)Aiming at the problem that the accuracy of calculating arm motion trajectory is not high with the traditional acceleration integral method,the attitude calculating position method is introduced,and the two methods are compared and analyzed through simulation experiments.The results show that the trajectory accuracy obtained by using the attitude calculating position method is higher,and the accuracy and stability of the tracking system are verified through the simple and complex movements of the arm.The simulation results show that the system can effectively track the arm movement.The results of this paper provide a reference for the research of robot teaching technology.