Growing user demand for computing power has spawned a large number of edge data center deployments

which have correspondingly increased energy costs and carbon emissions. In order to solve above problems

this paper presents an energy optimization management considering the coupling characteristics of computing power and electrothermal behaviors in edge data centers. Firstly

a collaborative energy management model is proposed

which integrates modules of task migration

battery power regulation

quality of service (QoS)

renewable energy consumption and power management. A task migration strategy based on internal Central Processing Unit temperature sensing and computing power adjustment is designed to avoid overheating and task blocking. Secondly

an optimization problem is formulated to minimize the long-term operating cost of edge data centers. Lastly

leveraging the Lyapunov optimization algorithm

the aforementioned issues are transformed into short-term online optimization problems. Simulation analysis demonstrates that the proposed method can effectively reduce the overall operational costs while ensuring long-term stable operation.