Research On Leg Rehabilitation Motion Capture System Based On Micro Inertial Sensors

A leg-rehabilitation motion capture system based on Micro-Electro-Mechanical System(MEMS) inertial sensor was designed. The aim of the design was to monitor patients’mo’vement and judge the fitness between patient’s movements and normal rehabilitation movements.The merits and drawbacks of motion capture systems available at market are discussed in this paper. MEMS inertial sensors has many characteristics. In consideration of the system structure design, working environment and the cost, the motion capture system based on MEMS inertial sensor has superiority than other system.Main contents of the study are as follows:(1) System structure include motion capture layer, data gathering layer and data transmit layer was designed. The wired connection between motion capture layer and data gathering layer was established through Controller Area Network(CAN). The wireless connection between data gathering layer and data transmit layer was established through ZIGBEE technique based on IEEE 802.15.4. It is authenticated that the system performs well in experiment. The network structure can complete the communication between the layers effectively.(2) The motion-capture-node was comprised of STMicrocontroller STM32F103T8U6,InvenSense MPU9150 and Texas Instruments VP230. The data-gather-node combined motion-capture-node and CC2530F256 to realize wireless communication. The data-transmit-node where the core was CC2531F256 established the connection between sensor network and computer. It is authenticated that the system can complete the acquisition,processing and transmission of motion signals.(3) In consideration of the features of MPU9150, a Finite Impulse Response(FIR) filter is designed to process the data collected by MPU9150. Magnetometer in MPU9150 is so sensitive that it is susceptible to external magnetic field interference. Ellipse fitting method is used to compensate the error of magnetometer. It is authenticated that the processed data improved in smoothness and stabilization.(4) After data-processing the 9-axes data is used to realize attitude calculation.Quaternion algorithm has a lot of advantages so it is used to calculate the attitude angle. The data collected by accelerator and magnetometer is used to calculate for initial attitude angle and get the initial quaternion by Gauss-Newton method. After that, a complementary filter and a Kalman filter is designed by using quaternion algorithm to realize real-time updates and real-time correction. It is authenticated that the the attitude angles updated with good accuracy in high-rate.After one-node and multi-node experiments, it is affirmed that the motion capture system can capture the movement of legs and the communication network is reliable. The calculated data can reproduce leg-movement and meet the requirement of exercise rehabilitation therapy.