Abstract: With the large-scale integration of distributed energy resources into power grids, microgrids face complex energy management challenges. The convergence of cloud-edge collaborative computing technology and artificial intelligence algorithms provides a new technical pathway for microgrid energy management. The system employs a three-tier collaborative architecture of cloud-edge-terminal, implements virtualized configuration of computing resources through Docker container technology, constructs a bi-level distributed energy scheduling model based on analytical target cascading, and designs intelligent scheduling algorithms for task migration and load balancing. Experimental validation demonstrates that the cloud-edge collaborative architecture exhibits excellent real-time performance and scalability in handling microgrid energy management tasks. Container virtualization technology shows significant advantages in resource utilization efficiency, providing an effective technical solution for constructing intelligent microgrid energy management systems.