Optimisation Strategies and Mathematical Modelling of the Path to Improvement of Students’ Physical Fitness Levels in Higher Education Physical Education

The development of physical education in colleges and universities can not only enrich the after-school lives of students and cultivate their sentiments, but also improve the physical fitness level of students and cultivate their sportsmanship in the new era. In this paper, we develop a diagnostic process for assessing the physical fitness level of college physical education students, present a dynamic monitoring framework for their physical fitness, and implement real-time data collection of students’ physical training using the LSTM model. The wavelet transform algorithm is used to get the dynamic features of the physical training data. It is then combined with the convolutional neural network to help the students recognize the features of their physical training. The associated dimension retrieval is then used to set up a test model for real-time monitoring and evaluation of the physical training data. The wavelet transform extracted the changing amplitude of the biceps brachii muscle, which ranged from -0.19 mV to 0.11 mV. The mean recognition accuracy of the students’ physical training features was 99.14%, and the RMSE fluctuation range of the students’ physical training monitoring was between 0.127 dB and 0.165 dB. The improvement of students’ physical fitness level in physical education in colleges and universities needs to start from the optimization of core strength and specialized physical fitness, which guides the promotion of physical fitness and the overall development of students in colleges and universities.