In order to study the model and system design of music dance resource management, a human nervous system-like fog computing architecture is proposed. Based on the Internet of Things, the music dance resource management system is analyzed, and the contribution model of fog computing resources and the allocation model of fog computing resources are established. The traditional NSGA-II (Quick Non-dominated Sorting Genetic Algorithms-II) algorithm is improved, and the improved NSGA-II algorithm is obtained to solve the resource allocation problem. Music dance resource management system is designed, implemented and tested. Through simulation and verification, the improved NSGA-II algorithm has better performance in average service delay time and average stability of task execution under different tasks and different fog nodes. In functional testing, the overall situation of music dance resource management system is good, and the design requirements are basically met. In performance testing, the response time is short and the throughput is high. Therefore, the system has high performance, and can meet the large amount of concurrent data access, which makes the user have better experience.