Athlete performance prediction based on time series analysis and implications for physical education and training strategies

In this paper, time series analysis is used to monitor and predict the performance of athletes in sports training.A smooth time series model , ARMA p q model is established, a fixed-order method based on autocorrelation function and partial correlation function is proposed, and the parameters of the model are estimated, and least squares prediction is used for model prediction.The monitoring test data of hemoglobin (HGB) in sports performance of Z athletes of a club were used as the research object, and the smooth time series test was conducted to determine the ARMA (1,1) model as the optimal time series fitting model, and the fitting effect was tested.In the application of blood oxygen saturation (BOS) index, ARMA (1,1) model can predict the trend of BOS of athlete Z with good application effect.Based on the prediction of athletes' performance by ARMA (1,1) model, this paper further proposes the integrated neuromuscular training method (INT), and integrates it with physical training will to develop the INT physical education training strategy.In the application experiment of INT physical education training strategy, the test results of the experimental group of athletes applying the INT physical education training strategy in the six events of T-test sensitive running, agility ladder, vestibular step, blindfolded one-legged standing, 30-meter sprint running, and 60-meter sprint running presented P<0.05, and the athletes' performance was significantly better than that of the control group.