Design and Implementation of a Universal Grid Computing Platform Applicable to Optimal Operation of Power Grid

Jialun Zhao, Libao Shi, Yixin Ni
DOI: https://doi.org/10.1049/cp.2019.0361
2019-01-01
Abstract:With the rapid development of the electric power industry, the demand for high performance computing is getting higher and higher. The advent and development of grid computing technology provides a strong guarantee in achieving highly efficient power grid simulations [1]. However, how to further improve the universality and usability of grid computing platform is always a difficult issue. This paper tries to design and construct a universal grid computing platform applicable to optimal operation of power grid with multi-task problemĀ­solving features. The designed platform can flexibly organize and schedule various heterogeneous computing resources by constructing a management node. In particular, the platform can identify newly added computing nodes automatically by receiving http requests and seamlessly integrate them into schedulable computing resources. An intelligent decisionĀ­making strategy, named "make the best use of everything", is proposed to realize the automatic division of tasks and dynamic balance of load, and the automatic conversion of different data formats can be efficiently implemented through the function of type recognition. Finally, the designed grid computing platform programming in Python is applied to implement the distributed parallel computation of optimal power flow solved by different computational intelligence algorithms based on specific test systems.
What problem does this paper attempt to address?