skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: On the stability and performance of discrete event methods for simulating continuous systems

Abstract

This paper establishes a link between the stability of a first order, explicit discrete event integration scheme and the stability criteria for the explicit Euler method. The paper begins by constructing a time-varying linear system with bounded inputs that is equivalent to the first order discrete event integration scheme. The stability of the discrete event system is shown to result from the fact that it automatically adjusts its time advance to lie below the limit set by the explicit Euler stability criteria. Moreover, because it is not necessary to update all integrators at this rate, a significant performance advantage is possible. Our results confirm and explain previously reported studies where it is demonstrated that a reduced number of updates can provide a significant performance advantage compared to fixed step methods. These results also throw some light on stability requirements for discrete event simulation of spatially extended systems.

Authors:
 [1];  [1]
  1. ORNL
Publication Date:
Research Org.:
Oak Ridge National Lab. (ORNL), Oak Ridge, TN (United States)
Sponsoring Org.:
USDOE Laboratory Directed Research and Development (LDRD) Program
OSTI Identifier:
931948
DOE Contract Number:
DE-AC05-00OR22725
Resource Type:
Journal Article
Resource Relation:
Journal Name: Journal of Computational Physics; Journal Volume: 227; Journal Issue: 1
Country of Publication:
United States
Language:
English
Subject:
99 GENERAL AND MISCELLANEOUS//MATHEMATICS, COMPUTING, AND INFORMATION SCIENCE; PERFORMANCE; SIMULATION; STABILITY

Citation Formats

Nutaro, James J, and Zeigler, Bernard P. On the stability and performance of discrete event methods for simulating continuous systems. United States: N. p., 2007. Web. doi:10.1016/j.jcp.2007.08.015.
Nutaro, James J, & Zeigler, Bernard P. On the stability and performance of discrete event methods for simulating continuous systems. United States. doi:10.1016/j.jcp.2007.08.015.
Nutaro, James J, and Zeigler, Bernard P. Mon . "On the stability and performance of discrete event methods for simulating continuous systems". United States. doi:10.1016/j.jcp.2007.08.015.
@article{osti_931948,
title = {On the stability and performance of discrete event methods for simulating continuous systems},
author = {Nutaro, James J and Zeigler, Bernard P},
abstractNote = {This paper establishes a link between the stability of a first order, explicit discrete event integration scheme and the stability criteria for the explicit Euler method. The paper begins by constructing a time-varying linear system with bounded inputs that is equivalent to the first order discrete event integration scheme. The stability of the discrete event system is shown to result from the fact that it automatically adjusts its time advance to lie below the limit set by the explicit Euler stability criteria. Moreover, because it is not necessary to update all integrators at this rate, a significant performance advantage is possible. Our results confirm and explain previously reported studies where it is demonstrated that a reduced number of updates can provide a significant performance advantage compared to fixed step methods. These results also throw some light on stability requirements for discrete event simulation of spatially extended systems.},
doi = {10.1016/j.jcp.2007.08.015},
journal = {Journal of Computational Physics},
number = 1,
volume = 227,
place = {United States},
year = {Mon Jan 01 00:00:00 EST 2007},
month = {Mon Jan 01 00:00:00 EST 2007}
}
  • The efficient and accurate management of time in simulations of hybrid models is an outstanding engineering problem. General a priori knowledge about the dynamic behavior of the hybrid system (i.e. essentially continuous, essentially discrete, or 'truly hybrid') facilitates this task. Indeed, for essentially discrete and essentially continuous systems, existing software packages can be conveniently used to perform quite sophisticated and satisfactory simulations. The situation is different for 'truly hybrid' systems, for which direct application of existing software packages results in a lengthy design process, cumbersome software assemblies, inaccurate results, or some combination of these independent of the designer's a priorimore » knowledge about the system's structure and behavior. The main goal of this paper is to provide a methodology whereby simulation designers can use a priori knowledge about the hybrid model's structure to build a straightforward, efficient, and accurate simulator with existing software packages. The proposed methodology is based on a formal decomposition and re-articulation of the hybrid system; this is the main theoretical result of the paper. To set the result in the right perspective, we briefly review the essentially continuous and essentially discrete approaches, which are illustrated with typical examples. Then we present our new, split system approach, first in a general formal context, then in three more specific guises that reflect the viewpoints of three main communities of hybrid system researchers and practitioners. For each of these variants we indicate an implementation path. Our approach is illustrated with an archetypal problem of power grid control.« less
  • From a modeling and simulation perspective, studying dynamic systems consists of focusing on changes in states. According to the precision of state changes, generic algorithms can be developed to track the activity of sub-systems. This paper aims at describing and applying this more natural and intuitive way to describe and implement dynamic systems. Activity is defined mathematically. A generic application case of diffusion is experimented with to compare the efficiency of quantized state methods using this new approach with traditional methods which do not focus computations on active areas. Our goal is to demonstrate that the concept of activity canmore » estimate the computational effort required by a quantized state method. Specifically, when properly designed, a discrete-event simulator for such a method achieves a reduction in the number of state transitions that more than compensates for the overhead it imposes.« less
  • Abstract not provided.
  • Abstract not provided.
  • Next-generation high-performance computing will require more scalable and flexible performance prediction tools to evaluate software--hardware co-design choices relevant to scientific applications and hardware architectures. Here, we present a new class of tools called application simulators—parameterized fast-running proxies of large-scale scientific applications using parallel discrete event simulation. Parameterized choices for the algorithmic method and hardware options provide a rich space for design exploration and allow us to quickly find well-performing software--hardware combinations. We demonstrate our approach with a TADSim simulator that models the temperature-accelerated dynamics (TAD) method, an algorithmically complex and parameter-rich member of the accelerated molecular dynamics (AMD) family ofmore » molecular dynamics methods. The essence of the TAD application is captured without the computational expense and resource usage of the full code. We accomplish this by identifying the time-intensive elements, quantifying algorithm steps in terms of those elements, abstracting them out, and replacing them by the passage of time. We use TADSim to quickly characterize the runtime performance and algorithmic behavior for the otherwise long-running simulation code. We extend TADSim to model algorithm extensions, such as speculative spawning of the compute-bound stages, and predict performance improvements without having to implement such a method. Validation against the actual TAD code shows close agreement for the evolution of an example physical system, a silver surface. Finally, focused parameter scans have allowed us to study algorithm parameter choices over far more scenarios than would be possible with the actual simulation. This has led to interesting performance-related insights and suggested extensions.« less