GPU-Accelerated Algorithm for On-Line Probabilistic Power Flow

Abstract

This letter proposes a superior GPU-accelerated algorithm for probabilistic power flow (PPF) based on Monte-Carlo simulation with simple random sampling (MCS-SRS). By means of offloading the tremendous computational burden to GPU, the algorithm can solve PPF in an extremely fast manner, two orders of magnitude faster in comparison to its CPU-based counterpart. Case studies on three large-scale systems show that the proposed algorithm can solve a whole PPF analysis with 10000 SRS and ultra-high-dimensional dependent uncertainty sources in seconds and therefore presents a highly promising solution for online PPF applications.

Department(s)

Electrical and Computer Engineering

Keywords and Phrases

Graphics Processing Unit; Intelligent Systems; Large Scale Systems; Monte Carlo Methods; Online Systems; Uncertainty Analysis; Case-Studies; Computational Burden; GPU-Accelerated; Online; Orders of Magnitude; Probabilistic Power Flow; Simple Random Sampling; Uncertainty Sources; Electric Load Flow; GPU; Monte-Carlo Simulation; Uncertainty Source

International Standard Serial Number (ISSN)

0885-8950

Document Type

Article - Journal

Document Version

Citation

File Type

text

Language(s)

English

Rights

© 2018 Institute of Electrical and Electronics Engineers (IEEE), All rights reserved.

Publication Date

01 Jan 2018

Link to Full Text

Share

 
COinS