Research on Intelligent Generation and Combination Optimization Methods for Service-Oriented APIs in Power Communication Data Centers
Zhongmiao Kang, Yuanjie Liu, Bo Li, Ying Zeng, Zhihua Alex Yang, Weilun Liu · 2025
With the digital and intelligent transformation of power systems accelerating, modern power communication networks are facing increasing challenges due to the heterogeneous, multi-source, and high-frequency nature of their data. Traditional API development in this domain is still largely manual, resulting in long development cycles and poor reusability. This paper proposes a novel intelligent mechanism for service-oriented API generation and combination optimization tailored for power communication data centers. The proposed approach includes a meta-model driven automatic API generation framework, a semantic-aware parameter inference module, and a reinforcement learning-based API combination optimization algorithm. A prototype platform was developed and deployed in a real-world scenario, demonstrating improvements in API generation efficiency (by over 65%) and performance in service routing. Experimental results verify the effectiveness and scalability of the proposed system, supporting agile service response in smart grid communication scenarios.