The traditional training system based on case teaching is according to the analysis of past competitions and training cases to carry out the strength training of aerobics special movements. The training results cannot be evaluated intelligently and accurately, and the performance of dynamic analysis is poor. To address this problem, the core training system of strength quality of aerobics special movements based on artificial intelligence is designed to realize the intelligent training of the strength quality of aerobics special movements. Through the study of fuzzy paradigm system, intelligent functions such as optimization and decision-making of intelligent fuzzy network are realized. The design system architecture framework includes the modules of sensor, receiver, database, and analysis decision. The core chip of the system is the main control module of Atmega1280 MCU for man-machine interaction, so as to realize the comprehensive training of the strength quality of aerobics special movements. Information collection module is used to collect information on strength training information such as instrument, movement, and language. The problem of phase distortion in signal transmission process is processed by FIR filter. Through information management module, trainee information management and training results statistics and queries are implemented. In the system software part, the system software structure diagram and system startup and landing procedure are given. By analyzing the working process of the module, the strength of aerobics special movements is analyzed. Experimental results show that the designed system can achieve real-time and stable strength training for aerobics special movements and improve training efficiency. [ABSTRACT FROM AUTHOR]