How to optimize course learning resources and learning methods to promote higher-order learning is an important research topic in current teaching reform. Based on the background of the big data era, this paper uses the MOOC network to build an open-sharing platform for physical education teaching course resources. Firstly, the MOOC network classroom teaching model is summarized through the analysis paradigm of the MOOC platform analyzing indicators to deal with teaching big data. The third-order nonlinear model is used in the teaching model to dynamically express the physical education teaching resource-sharing design system and establish a physical education simulation model. In the simulation model, the search optimal solution is improved for resource platform search, and global search is performed when the inertia factor is large and local search is performed when it is small. The univariate validation test for group S students revealed that the mean value of the performance of the answers selected by group S students before the test compared to after the test changed from negative to positive, from -0.46 to 0.16, which was a significant improvement. It can be seen that the teaching model based on the MOOC web-supported open-sharing platform of physical education course resources has positively contributed to students’ learning performance, learning interest, and independent learning ability.