The physiological signals of badminton players vary greatly in different contexts, with obvious diversity and complexity. In the emotion recognition, the recognition cycle affects the recognition interval, which in turn leads to abnormal emotion recognition of physiological signals. Therefore, a new emotion recognition method for badminton players' physiological signals is designed based on hierarchical recurrent genetic algorithm. That is, we extracted the emotion recognition features of badminton players’ physiological signals, searched for the emotion recognition of badminton players' physiological signals using hierarchical recurrent genetic algorithm, and designed an emotion recognition algorithm of badminton players' physiological signals. This process enabled us to successfully identify and interpret the emotional cues present in the physiological signals of badminton players. The experimental results show that the designed method based on hierarchical recurrent genetic algorithm for badminton players' physiological signal emotion recognition does not have abnormal recognition problems, and can normally recognize physiological signals emotion recognition in different situations, which proves that the designed method of physiological signal emotion recognition has a better recognition effect, is reliable, and has a certain value of application, and makes certain contribution to the development of reasonable training strategy and timely emergency disposal.