Based on the HTTP adaptive streaming transmission scheme, it can cope with complex and ever-changing network environments and reception conditions, effectively absorbing the advantages of progressive and real-time, and solving their respective problems. The learning system based on video on demand learns under various sensory stimuli, improving learning efficiency and interest. This article first studies a streaming media transmission system composed of servers, networks, and clients; Then, the DASH system architecture consisting of slice parsing module, client control module, information monitoring module, MPD parsing module, and media player was studied; Next, design adaptive algorithms based on DASH, including bandwidth estimation, bit rate selection, and cache control; Finally, simulation experiments were conducted to verify the superiority of the adaptive streaming algorithm for live streaming systems based on the DASH architecture. The research results fully leverage the advantages of DASH, supporting a wider range of media sharding formats and dynamic setting of sharding times.