A Faster Non-Linear Iteration Solver using Graph Computing and its Application in Power Flow Calculation. (December 2017)

• Record Type: Journal Article
• Authors: Liu, Kewen; Liu, Guangyi; Xie, Kai; Wang, Zhiwei
• Is Part Of: Energy procedia. Volume 142(2017)
• Page Start: 2534
• Page End: 2540
• Publication Date: 2017-12
• Subjects: Graph theory -- graph database -- graph computing -- power system analysis -- graph data modeling -- PageRank Algorithm -- Bulk Synchronous Parallel Computing Model; Power resources -- Congresses; Power resources -- Periodicals; Power resources; Conference proceedings; Periodicals; 333.7905
• DOI: 10.1016/j.egypro.2017.12.194
• Languages: English
• ISSNs: 1876-6102

Abstract

Abstract: To solve large scale fast matrix calculation problems, a Non-Linear Iteration Solver (NLIS) based on a BSP model using a graph database & computing technology is proposed in this paper. A node and edge level parallel iterative algorithm is implemented by using a graph computing algorithm combined with a message passing mechanism. IEEE testing cases and actual operation data of a provincial power system in China are used to verify the method, then expand the MP 10790 to over 100k and 1 million nodes systems for the calculation stress test on a conventional server.