Optimization of intelligent physical education teaching method based on fuzzy logic and virtual reality

Traditional physical education classes frequently face challenges with student participation, resource availability, and individualized instruction, despite the importance of physical education to overall health, well-being, and physical fitness. Utilizing Virtual Reality (VR)–augmented settings, Salp Swarm Optimization–Seq2Seq Neural Networks (SSO-NN), and fuzzy logic adaptation, this research presents a novel automated human activity detection system designed to address the issues presented. To obtain spectral density, skewness, and kurtosis, multivariate time-series motion data from wearable sensors, such as accelerometers, are studied. Salp Swarm Optimization selects the most informative temporal variables and models them with a Sequence-to-Sequence neural network to recognize human activity sequences of unknown length. Fuzzy logic enables adjusting difficulty and tailoring instruction based on the student’s level of performance and involvement. It has been demonstrated using real-world physical activity datasets that the proposed system can recognize complex movements with high accuracy and provides customizable, adaptive teaching. This effort achieved primary improvements in virtual reality (VR) immersion, enhanced activity recognition, and fuzzy reasoning. These innovations enhance student engagement and enable the customization of physical education curricula, thereby improving the overall learning experience.