Title: Quasi-parallel network applications in real-time distribution management system
Authors: Izudin Dzafic; Hans-Theo Neisius
Addresses: Siemens AG, Humboldtstr 50, D-90459, Nuremberg, Germany. ' Siemens AG, Humboldtstr 50, D-90459, Nuremberg, Germany
Abstract: In modern distribution management systems (DMSs), the real-time calculation of the network's state plays an important role for the operation of the network. High sophisticated algorithms for the estimation and calculation of the system's state like distribution system state estimation (DSSE), distribution system power and short circuit calculations (DSPF and SCC), voltage var control (VVC) and others must operate in a real-time environment, supplying subsequently running functions with reliable input data. Closed-loop requirements also increase the requirements concerning performance, availability and robustness. In this paper, an implementation of the sequence control of different DMS-network applications is presented. This proposed real-time scheduler takes advantage of CPU-cores inside of a modern computer system: by passing subtasks to different CPU-cores, the overall performance is significantly increased. The scheduling process is analysed in detail taking the nature of a distribution system into consideration. The presentation is finalised with a report about the gain of performance for the state estimation process.
Keywords: distribution management systems; DMS; multiprocessors; real-time distribution; scheduling; closed loop; smart grid; sequence control; state estimation.
DOI: 10.1504/IJICA.2012.045707
International Journal of Innovative Computing and Applications, 2012 Vol.4 No.1, pp.3 - 11
Received: 13 Oct 2011
Accepted: 29 Nov 2011
Published online: 17 Sep 2014 *