Background: Nowadays, people are under a lot of stress in life, and they often seek some ways to release the pressure, and ping shuai gong is one of them. However, because there are no objective indicators, beginners cannot know whether there is a relaxing effect and increase the difficulty of promotion. Therefore, this study hopes to develop a simple and lightweight brain wave to bring objective and quantitative effects on the performance of ping shuai gong. Problems and Hypothesis: The brain wave band developed in this research can provide objective quantitative indicators through machine learning to analyze the user's brain wave. Materials and methods: Use different dry electrodes to develop the brain wave band and test the signal quality. The subject was guided by a special person to measure the level swing function and wear the brain wave band developed in this study to measure the brain wave (FpZ & Oz) when the level swing function was measured. The experimental procedure is as follows: first record the basic value for 5 minutes and open the eyes for 2 minutes Half, with eyes closed for 2 and a half minutes, start to ping shuai gong for 30 minutes, and then finish for 2 minutes and a half for eyes closed and 2 minutes and half for the eyes. The total experimental time is 40 minutes. The measured ping shuai gong data was divided into inexperienced and experienced persons. Analyze the brain wave frequency spectrum, and then use artificial neural networks (Artificial Neural Networks, ANN) and convolutional neural networks (Convolutional Neural Network, CNN) to distinguish the effectiveness of the subjects. Preliminary results: The accuracy of CNN model classification of those who have experience in ping shuai gong and those who have no experience can reach 90%, but there is a phenomenon of overfitting. Importance:Establish a convenient, objective, and accurate brainwave belt physiological feedback system. Users can use this system to perform self-level training.
