This study develops and validates an intervention system for college students' tennis training. From the perspective of mindfulness meditation, it integrates virtual reality (VR) technology and artificial intelligence (AI) through a Back Propagation Neural Network (BPNN) model. The experiment divides respondents into three groups: VR mindfulness group, control group, and mindfulness group. It systematically evaluates changes in training performance, sleep quality, emotional states, fatigue perception, and mindfulness levels while constructing a BPNN model to predict training outcomes. Results demonstrate that the VR mindfulness group achieves the most remarkable performance improvement (average 9.53%), substantially exceeding the mindfulness group (8.13%) and control group (4.31%), with all p-values <0.05. Psychological indicators, including mindfulness levels, fatigue perception, emotional states, and sleep quality, also exhibit marked improvement trends. The BPNN model shows strong predictive capability, with all three groups maintaining mean absolute percentage errors below 1%. Further analysis of the intermediary roles reveals that sleep quality, fatigue perception, mindfulness levels, and emotional states are all significant mediating paths. In contrast, the indirect path of mindfulness levels between training methods and performance is the most prominent, with a p-value<0.001. The study indicates that AI-enhanced VR mindfulness training effectively boosts college students' tennis training outcomes. Moreover, it achieves indirect promotion by optimizing the psychological state, which holds substantial and theoretical value and provides an empirical basis for intelligent sports training.
Liao et al. (Fri,) studied this question.