With the rapid increase in business types and quantities such as unmanned driving and smart factories, the importance of computing power network resources is becoming increasingly high. To solve the problem of low utilization of computing power network resources, this paper takes the computing power network as the service function chain to allocate resources as the research object, and designs a resource allocation algorithm based on DDQN in the computing power network environment. Firstly, a network model was constructed for the research problem from two dimensions: computing power network and service function chain. Secondly, the network model was modeled as a Markov decision process, and the action network and target network environment were constructed. Finally, a resource allocation algorithm based on DDQN was designed. This algorithm prioritizes resource allocation for initiated computing power nodes when selecting actions. During performance analysis, it was experimentally verified that the algorithm proposed in this paper has achieved good results in improving resource utilization.