A Global Data Model for Electric Power Data Centers

摘要

With the development of large-scale regional interconnected power grid, smart electronic devices and phasor measurement units have been widely used. How to realize high-speed processing of global data in power enterprise data center has become the key to real-time computing. This paper presents a data center global data model (DCGDM) for power enterprises. According to the computational coupling between tasks, the global data model of data center is established. With the constraints of CPU utilization and memory utilization of virtual machine and the goal of energy saving, the optimal configuration of the global data model of power enterprise data center is established. The results show that DCGDM is superior to traditional single-machine multi-threaded parallel computing in terms of time and system consumption. The calculation time of IEEE118 nodes is shortened by 42.32%, and the actual calculation time of 1133 nodes is shortened by 75.8% with the increase of system scale.