Efficient Simulation of Physical System Models Using Inlined Implicit Runge-Kutta Algorithms Vicha Treeaporn Department of Electrical & Computer Engineering.

Slides:



Advertisements
Similar presentations
Formal Computational Skills
Advertisements

Alexei Medovikov Tulane University
Chapter 6 Differential Equations
Partial Differential Equations
Computational Modeling for Engineering MECN 6040
Integration Techniques
Ordinary Differential Equations
Error Measurement and Iterative Methods
On Stiffness in Affine Asset Pricing Models By Shirley J. Huang and Jun Yu University of Auckland & Singapore Management University.
Total Recall Math, Part 2 Ordinary diff. equations First order ODE, one boundary/initial condition: Second order ODE.
CSCE Review—Fortran. CSCE Review—I/O Patterns: Read until a sentinel value is found Read n, then read n things Read until EOF encountered.
CSE245: Computer-Aided Circuit Simulation and Verification Lecture Note 5 Numerical Integration Spring 2010 Prof. Chung-Kuan Cheng 1.
AE/ME 339 Computational Fluid Dynamics (CFD) K. M. Isaac Professor of Aerospace Engineering.
Chapter 1 Introduction The solutions of engineering problems can be obtained using analytical methods or numerical methods. Analytical differentiation.
François E. Cellier and Matthias Krebs
Development of Empirical Models From Process Data
Harvard University - Boston University - University of Maryland Numerical Micromagnetics Xiaobo TanJohn S. Baras P. S. Krishnaprasad University of Maryland.
Ordinary Differential Equations (ODEs) 1Daniel Baur / Numerical Methods for Chemical Engineers / Implicit ODE Solvers Daniel Baur ETH Zurich, Institut.
MCE 561 Computational Methods in Solid Mechanics
Digital Filter Stepsize Control in DASPK and its Effect on Control Optimization Performance Kirsten Meeker University of California, Santa Barbara.
Ordinary Differential Equations (ODEs)
MATH 685/ CSI 700/ OR 682 Lecture Notes Lecture 10. Ordinary differential equations. Initial value problems.
P ROJECT : N UMERICAL S OLUTIONS TO O RDINARY D IFFERENTIAL E QUATIONS IN H ARDWARE Joseph Schneider EE 800 March 30, 2010.
Ordinary Differential Equations (ODEs) 1Daniel Baur / Numerical Methods for Chemical Engineers / Explicit ODE Solvers Daniel Baur ETH Zurich, Institut.
Numerical Solutions to ODEs Nancy Griffeth January 14, 2014 Funding for this workshop was provided by the program “Computational Modeling and Analysis.
Professor Walter W. Olson Department of Mechanical, Industrial and Manufacturing Engineering University of Toledo Solving ODE.
Boyce/DiPrima 9th ed, Ch 8.4: Multistep Methods Elementary Differential Equations and Boundary Value Problems, 9th edition, by William E. Boyce and Richard.
Lecture 35 Numerical Analysis. Chapter 7 Ordinary Differential Equations.
Polynomial Chaos For Dynamical Systems Anatoly Zlotnik, Case Western Reserve University Mohamed Jardak, Florida State University.
Algorithms for a large sparse nonlinear eigenvalue problem Yusaku Yamamoto Dept. of Computational Science & Engineering Nagoya University.
Chapter 17 Boundary Value Problems. Standard Form of Two-Point Boundary Value Problem In total, there are n 1 +n 2 =N boundary conditions.
EE3561_Unit 8Al-Dhaifallah14351 EE 3561 : Computational Methods Unit 8 Solution of Ordinary Differential Equations Lesson 3: Midpoint and Heun’s Predictor.
1 EEE 431 Computational Methods in Electrodynamics Lecture 4 By Dr. Rasime Uyguroglu
Increasing asymptotic stability of Crank-Nicolson method Alexei A. Medovikov Vyacheslav I. Lebedev.
Modelling & Simulation of Chemical Engineering Systems Department of Chemical Engineering King Saud University 501 هعم : تمثيل الأنظمة الهندسية على الحاسب.
Integration of 3-body encounter. Figure taken from
Professor Walter W. Olson Department of Mechanical, Industrial and Manufacturing Engineering University of Toledo Performance & Stability Analysis.
Major objective of this course is: Design and analysis of modern algorithms Different variants Accuracy Efficiency Comparing efficiencies Motivation thinking.
Separation of Variables Solving First Order Differential Equations.
Copyright © 2006 The McGraw-Hill Companies, Inc. Permission required for reproduction or display. ~ Ordinary Differential Equations ~ Stiffness and Multistep.
On the Use of Sparse Direct Solver in a Projection Method for Generalized Eigenvalue Problems Using Numerical Integration Takamitsu Watanabe and Yusaku.
Chapter 3 Roots of Equations. Objectives Understanding what roots problems are and where they occur in engineering and science Knowing how to determine.
Rieben IMA Poster, 05/11/ UC Davis /LLNL/ ISCR High Order Symplectic Integration Methods for Finite Element Solutions to Time Dependent Maxwell Equations.
THE LAPLACE TRANSFORM LEARNING GOALS Definition
ECE 576 – Power System Dynamics and Stability Prof. Tom Overbye Dept. of Electrical and Computer Engineering University of Illinois at Urbana-Champaign.
MECN 3500 Inter - Bayamon Lecture 9 Numerical Methods for Engineering MECN 3500 Professor: Dr. Omar E. Meza Castillo
Large Timestep Issues Lecture 12 Alessandra Nardi Thanks to Prof. Sangiovanni, Prof. Newton, Prof. White, Deepak Ramaswamy, Michal Rewienski, and Karen.
MECH4450 Introduction to Finite Element Methods Chapter 9 Advanced Topics II - Nonlinear Problems Error and Convergence.
Copyright © The McGraw-Hill Companies, Inc. Permission required for reproduction or display. 1 Part 7 - Chapter 25.
Announcements Read Chapters 11 and 12 (sections 12.1 to 12.3)
Final Project Topics Numerical Methods for PDEs Spring 2007 Jim E. Jones.
ME451 Kinematics and Dynamics of Machine Systems Numerical Integration. Stiffness: Implicit Methods. October 30, 2013 Radu Serban University of Wisconsin-Madison.
ECE 530 – Analysis Techniques for Large-Scale Electrical Systems
ECE 530 – Analysis Techniques for Large-Scale Electrical Systems Prof. Hao Zhu Dept. of Electrical and Computer Engineering University of Illinois at Urbana-Champaign.
ECE 576 – Power System Dynamics and Stability Prof. Tom Overbye Dept. of Electrical and Computer Engineering University of Illinois at Urbana-Champaign.
Game Technology Animation V Generate motion of objects using numerical simulation methods Physically Based Animation.
Population Based Optimization for Variable Operating Points Alan L. Jennings & Ra úl Ordóñez, ajennings1ajennings1,
Ordinary Differential Equations
Sec 21: Generalizations of the Euler Method Consider a differential equation n = 10 estimate x = 0.5 n = 10 estimate x =50 Initial Value Problem Euler.
Circuit Simulation using Matrix Exponential Method Shih-Hung Weng, Quan Chen and Chung-Kuan Cheng CSE Department, UC San Diego, CA Contact:
Copyright © The McGraw-Hill Companies, Inc. Permission required for reproduction or display. 1 Part 6 - Chapters 22 and 23.
VEHICLE DYNAMICS SIMULATIONS USING NUMERICAL METHODS VIYAT JHAVERI.
Computational Fluid Dynamics Lecture II Numerical Methods and Criteria for CFD Dr. Ugur GUVEN Professor of Aerospace Engineering.
ECE 576 – Power System Dynamics and Stability
Chapter 2 Interconnect Analysis
Class Notes 19: Numerical Methods (2/2)
CSE 245: Computer Aided Circuit Simulation and Verification
4th Homework Problem In this homework problem, we wish to exercise the tearing and relaxation methods by means of a slightly larger problem than that presented.
Overview Class #2 (Jan 16) Brief introduction to time-stepping ODEs
Presentation transcript:

Efficient Simulation of Physical System Models Using Inlined Implicit Runge-Kutta Algorithms Vicha Treeaporn Department of Electrical & Computer Engineering The University of Arizona Tucson, Arizona U.S.A

Topics Introduction Introduction Techniques for Simulation Techniques for Simulation Results Results An Application An Application

Introduction Stiffness Stiffness Widely varying eigenvalues Widely varying eigenvalues Explicit algorithms Explicit algorithms Straightforward to implement Straightforward to implement Step size limited by numerical stability Step size limited by numerical stability Implicit algorithms Implicit algorithms More difficult to implement More difficult to implement Additional computational load Additional computational load Needed to simulate stiff systems Needed to simulate stiff systems May use larger step sizes May use larger step sizes

Inline-Integration Merges the integration algorithm with the model Merges the integration algorithm with the model Eliminates differential equations Eliminates differential equations Results in difference equations (∆Es) Results in difference equations (∆Es) Easily implement implicit algorithms Easily implement implicit algorithms Circuit example inlining Rad3 Circuit example inlining Rad3

Simple Circuit

Circuit Equations

Inlined with Rad3 Integrator equations Eliminate derivatives Evaluate at Rad3 time instants

Sorting

Sorting

Sorting

Sorting 10 equations immediately causalized 10 equations immediately causalized Need to perform tearing Need to perform tearing Make assumptions about variables being ‘known’ Make assumptions about variables being ‘known’

Tearing Residual Eq. Tearing variable

Tearing Residual Eq. #2 Tearing variable #2

Tearing Completely causalized equations Completely causalized equations 2 iteration variables, v c and i 1 2 iteration variables, v c and i 1 Could use this set of equations for simulation Could use this set of equations for simulation Want step-size control Want step-size control

Step-Size Control Want larger step sizes Want larger step sizes Reduce the overall computational cost Reduce the overall computational cost Maintain desired accuracy Maintain desired accuracy Compute error estimate Compute error estimate Embedding method Embedding method Shares computations with original method Shares computations with original method

Step-Size Control Explicit RKs Explicit RKs Embedding methods have been found Embedding methods have been found Implicit RKs Implicit RKs Difficult problem Difficult problem Algorithms are compact Algorithms are compact Can find embedding methods using two steps Can find embedding methods using two steps Linear polynomial approximation Linear polynomial approximation

HW-SDIRK Embedding 3 rd -order accurate 3 rd -order accurate Behaves like an explicit method Behaves like an explicit method May unnecessarily restrict step size for stiff systems May unnecessarily restrict step size for stiff systems Search for an alternate embedding method Search for an alternate embedding method

Alt. HW-SDIRK Embedding 3 rd -order accurate 3 rd -order accurate Implicit method Implicit method

Alt. HW-SDIRK Embedding Stability Domain Damping Plots

Lobatto IIIC(6) No embedding method exists No embedding method exists Expensive to perform step size control Expensive to perform step size control Can search for an embedding method Can search for an embedding method

Lobatto IIIC(6) Embedding Method 5 th -order accurate 5 th -order accurate A-Stable A-Stable Large asymptotic region Large asymptotic region

Lobatto IIIC(6) Embedding Method Stability Domain Damping Plots

Numerical Experiments

Tested various algorithms with selected benchmark ODEs Tested various algorithms with selected benchmark ODEs Implemented in Dymola/Modelica Implemented in Dymola/Modelica

ODE Set B ode15s Inlined with HWSDIRK and alternate error method

ODE Set B Error estimate stays near Step size grows and shrinks appropriately

ODE Set D Inlined with Lobatto IIIC(6) ode15s

ODE Set D

An Application

Real-Time, Limited Resources Real-Time, Limited Resources Embedded control systems Embedded control systems Model Predictive Model Predictive Add additional system dynamics Add additional system dynamics Simulate missile dynamics in flight for trajectory shaping Simulate missile dynamics in flight for trajectory shaping First solution is faster computer First solution is faster computer Model may still be too complex Model may still be too complex Try inlining Try inlining

Questions?