Martial Arts Training Prediction Model Based on Big Data and MEMS Sensors

In martial arts teaching and sports training, the accurate capturing analysis of athletes’ posture is conducive to accurately judge postures, as well correcting movements in a targeted manner, further improving performance reducing physical damage. The manufacturing level MEMS sensors continues improve, status perception assembly objects becoming more abundant accurate. shape small can be worn, data collected continuously without obstacles. price relatively low, privacy protection strong, advantages are clear prominent. A considerable number technicians choose use main tool for human behavior detection collection. Therefore, this article designs multiple inertial form body lower limb capture device, its core components composed accelerometer, gyroscope, magnetometer. order make obtained acceleration value, angular velocity magnetometer value reflect movement state structure, different fusion algorithms ellipsoid fitting calibration studied realize calculation angle each joint point obtain big data. addition, through analysis, training injury risk prediction model, which provide guidance suggestions tasks.