A Study of Data-Driven Optimization of Personalized Instructional Strategies for Physical Education Training

Abstract

As a new type of sports training, the application of sports information technology can effectively improve the athletes’ sports level and sports performance. In this paper, based on the way and effect of sports information technology application, a system for sports teaching and training assistance is designed. Through sports training intensity monitoring, the training intensity of the athletes is obtained, and a high load alarm is triggered. At the same time, the improved hybrid Kalman filter is used to solve the posture of training movements, and the DTW algorithm is used to recognize the training movements and compare them with the standard movements to obtain the training scores. Coaches can be assisted in formulating personalized teaching strategies by using the system’s obtained sports training data. This paper’s sports teaching and training assistance system has a significantly shorter response time than the two comparative systems, which suggests it has better response performance and practicality. The RMSE values of the improved hybrid Kalman filter are 0.92, 0.84 and 1.47 for the roll, pitch and yaw angles, respectively, which are better than that of the extended Kalman filter, indicating that it is closer to the original data and has better data fusion performance. Compared with the control class using traditional teaching strategies, the students in the experimental class using the personalized teaching strategies of this paper have higher skill improvement, indicating that the strategies of this paper have an important guiding role for personalized physical education teaching.

Keywords

• extended kalman filter
• pose solving
• dtw algorithm
• training movement recognition
• physical education teaching
• 97m80