Design of power data center based on Dempster-Shafer and wireless sensor network sharing

Design of power data center based on Dempster-Shafer and wireless sensor network sharing

연구 분야: Networking

학회: Discover Computing

In this paper, the design of a power data center based on dempster-shafer and wireless sensor network sharing is proposed to solve the problems of poor intelligent scheduling ability and long transmission delay in traditional power information management systems. By constructing a wireless sensor information collection model of power big data under the smart power management microservice architecture, combined with the wireless sensor network sharing and scheduling scheme, an intelligent deployment routing algorithm for network nodes is established. The power data center prediction and feature extraction under the wireless sensor network sharing and fusion mechanism are realized using Dempster-Shafer evidence theory system, and an intelligent power management microservice platform based on the design of power data center is constructed. The test results show that the corresponding value of this method is generally higher than that of the reference in the comparison of the balanced dispatching output. The transmission delay of this method is reduced by 43.20% and 52.00%, respectively, compared with the average transmission delay of references (Wankai et al. in Power Syst Technol 46:4473–4481, 2022) and (Changyue et al. in IEEE Trans Power Electron 33:10157–10168, 2018). These results show that the method in this study has a significant effect on reducing the transmission delay in the design of a power data center. This method can improve the fusion level and intelligent scheduling ability of power data, improve data clustering and stability of power data transmission, and improve the sharing level of power data. It is proven that the power data center and information management system under the intelligent power management micro-service framework constructed by this method have improved the power service level. The analysis shows that this method has a small delay in power data output and a high stability in information collection and transmission.