CloudPSS: A high-performance power system simulator based on cloud computing
Yankan Song, Ying Chen, Zhitong Yu, Shaowei Huang, Chen Shen
Abstract
With the increasing computations in power system simulations, high-performance and cost-effective power system simulator is highly required. In this paper, a cloud-computing based power system simulator, namely CloudPSS, is designed. Based on an open service integrating framework, a self-developed electromagnetic transients (EMT) simulator with an automatic code generator is provided to accelerate EMT simulations using heterogeneous computing devices in the cloud, such as CPU and GPU. Test results show that the CloudPSS platform can achieve significant speedups for both large-scale simulation tasks and multi-scenario tasks. Moreover, benefit from the time-sharing of cloud computing resources, the computational cost can be greatly reduced.