A Supernodal Approach to Sparse Partial Pivoting
SIAM Journal on Matrix Analysis and Applications
LAPACK Users' guide (third ed.)
LAPACK Users' guide (third ed.)
A Fully Asynchronous Multifrontal Solver Using Distributed Dynamic Scheduling
SIAM Journal on Matrix Analysis and Applications
Reducing the bandwidth of sparse symmetric matrices
ACM '69 Proceedings of the 1969 24th national conference
Continuous System Simulation
BCYCLIC: A parallel block tridiagonal matrix cyclic solver
Journal of Computational Physics
Decentralized agent-based underfrequency load shedding
Integrated Computer-Aided Engineering - Multi-Agent Systems for Energy Management
| Hi-index | 0.00 |
Future electric grid technology is envisioned on the notion of a smart grid in which responsive end-user devices play an integral part of the transmission and distribution control systems. Detailed simulation is often the primary choice in analyzing small network designs, and the only choice in analyzing large-scale electric network designs. Here, we identify and articulate the high-performance computing needs underlying high-resolution discrete event simulation of smart electric grid operation large network scenarios such as the entire Eastern Interconnect. We focus on the simulator's most computationally intensive operation, namely, the dynamic numerical solution for the electric grid state, for both time-integration as well as event-detection. We explore solution approaches using general-purpose dense and sparse solvers, and propose a scalable solver specialized for the sparse structures of actual electric networks. Based on experiments with an implementation in the THYME simulator, we identify performance issues and possible solution approaches for smart grid experimentation in the large.