La traduction de cette page n'est pas à jour. Cliquez ici pour voir la dernière version en anglais.

Configurer les conditions de simulation

Sélectionner un solveur, définir les conditions initiales, choisir un jeu de données d'entrée et spécifier la taille de pas

Après avoir créé un modèle dans Simulink^®, vous pouvez configurer la simulation pour qu'elle s'exécute de manière rapide et précise sans apporter de modifications structurelles au modèle.

La première étape de configuration de la simulation est la sélection d'un solveur. Par défaut, Simulink sélectionne automatiquement un solveur à pas variable. Vous pouvez ajuster les options du solveur ou en sélectionner un autre dans le volet « Solver » de la boîte de dialogue « Configuration Parameters ».

Il peut arriver qu'une simulation ralentisse ou se bloque. Utilisez Solver Profiler pour identifier les goulots d'étranglement de la simulation et obtenir des recommandations afin d'améliorer la performance du solveur.

Fonctions

développer tout

Jeu de configuration

`openDialog`	Open configuration parameters dialog
`closeDialog`	Close configuration parameters dialog

Diagnostics

`Simulink.BlockDiagram.getAlgebraicLoops`	Identify and analyze algebraic loops in a model
`solverprofiler.profileModel`	Programmatically analyze solver performance for model using Solver Profiler

Outils

développer tout

Solver Profiler

Zero Crossing Explorer

State Explorer

Paramètres du modèle

développer tout

Durée de la simulation et sélection du solveur

Start time	Simulation start time
Stop time	Simulation stop time
Type	Choice of variable- or fixed-step solver
Solver	Solver that computes states and outputs for simulation

Paramètres du solveur à pas variable

Taille de pas et tolérance d'erreur

Max step size	Maximum step size for variable-step solver
Min step size	Minimum step size for variable-step solver
Initial step size	Size of first time step for variable-step solver
Number of consecutive min steps	Number of steps less than or equal to minimum step size allowed before minimum step size violation occurs
Relative tolerance	Relative tolerance for solver tolerance calculation
Absolute tolerance	Absolute tolerance for solver tolerance calculation
Auto scale absolute tolerance	Option to scale absolute tolerance based on state values
Shape preservation	Option to preserve shape of states using derivative information at each time step

Détection des passages par zéro

Zero-crossing control	Option to control how zero-crossing detection is enabled in the model
Algorithm	Algorithm for zero-crossing detection with variable-step solver
Time tolerance	Definition of consecutive zero crossings
Signal threshold	State value at which adaptive zero-crossing algorithm can stop bracketing
Number of consecutive zero crossings	Threshold for issuing diagnostic due to consecutive zero crossings

Options de calcul du solveur

Integration method	Integration for nonadaptive `odeN` variable-step solver (depuis R2020a)
Maximum order	Order of numerical differentiation formulas used for `ode15s` solver
Solver reset method	Option to specify whether solver recomputes Jacobian matrix during solver reset
Solver Jacobian Method	Method implicit solvers use to compute Jacobian matrix
Extrapolation order	Extrapolation order for `ode14x` fixed-step solver
Number of Newton's iterations	Number of Newton's method iterations used by `ode14x` and `ode1be` solvers
Daessc mode	Mode of operation for `daessc` solver

Options de gestion des tâches et des pas d'échantillonnage

Automatically handle rate transition for data transfer	Option to ensure integrity of data transfer between different sample times in deployed code
Allow multiple tasks to access inputs and outputs	Option to treat root-level input and output ports as part of each connected task in rate-based model (depuis R2021b)
Higher priority value indicates higher task priority	Priority ordering for real-time system targets
Deterministic data transfer	Deterministic data transfer behavior for automatically inserted Rate Transition blocks

Paramètres du solveur à pas fixe

Taille de pas

Fixed-step size (fundamental sample time)

Step size for fixed-step solver

Détection des passages par zéro

Enable zero-crossing detection for fixed-step solver	Option to use zero-crossing detection with fixed-step solver (depuis R2022a)
Zero-crossing control	Option to control how zero-crossing detection is enabled in the model
Maximum number of bracketing iterations	Maximum number of iterations performed when locating zero crossing (depuis R2022a)
Maximum number of zero-crossings per step	Maximum number of zero crossings to locate in a single time step (depuis R2022a)

Options de calcul du solveur

Solver Jacobian Method	Method implicit solvers use to compute Jacobian matrix
Extrapolation order	Extrapolation order for `ode14x` fixed-step solver
Number of Newton's iterations	Number of Newton's method iterations used by `ode14x` and `ode1be` solvers

Options de gestion des tâches et des pas d'échantillonnage

Periodic sample time constraint	Option to specify constraints on model sample times
Sample time properties	Discrete sample time periods, offsets, and priorities
Treat each discrete rate as a separate task	Option to enable multitasking execution
Allow tasks to execute concurrently on target	Enable concurrent tasking behavior for model
Automatically handle rate transition for data transfer	Option to ensure integrity of data transfer between different sample times in deployed code
Allow multiple tasks to access inputs and outputs	Option to treat root-level input and output ports as part of each connected task in rate-based model (depuis R2021b)
Higher priority value indicates higher task priority	Priority ordering for real-time system targets
Deterministic data transfer	Deterministic data transfer behavior for automatically inserted Rate Transition blocks

Comportement de diagnostic du solveur

Algebraic loop	Diagnostic behavior when algebraic loop detected during compilation
Minimize algebraic loop	Diagnostic behavior when the software is unable to resolve artificial algebraic loops
Block priority violation	Diagnostic behavior when the software detects block priority specification error
Min step size violation	Diagnostic behavior when minimum step size violation occurs
Consecutive zero-crossings violation	Diagnostic behavior when zero-crossing violation occurs
Automatic solver parameter selection	Diagnostic behavior when the software changes a solver parameter value
State name clash	Diagnostic behavior when more than one state has same name

Rubriques

Solveurs dans Simulink

Choose a Solver

Choose a solver based on the dynamics of the model.
- Variable Step Solvers in Simulink
- Fixed Step Solvers in Simulink
- Choose a Fixed-Step Solver
Compare Solvers
A dynamic system is simulated by computing its states at successive time steps over a specified time span, using information provided by the model.
Zero-Crossing Detection
Learn how zero-crossing events affect simulation.
Choose a Jacobian Method for an Implicit Solver
For implicit solvers, Simulink must compute the solver Jacobian, which is a submatrix of the Jacobian matrix associated with the continuous representation of a Simulink model.
Use Local Solvers in Referenced Models
Learn how local solvers work and how to configure a local solver.
Zero-Crossing Detection with Fixed-Step Simulation
Learn how zero-crossing detection affects fixed-step simulation.

Informations d'état

Save Block States and Simulation Operating Points
Learn how you can use state information logged from simulation and decide how to log states and operating points.
Use Model Operating Point for Faster Simulation Workflow
Reduce the time required to run a set of simulations by simulating from an initial operating point.

Solver Profiler

Examine Model Dynamics Using Solver Profiler
Identify factors affecting model simulation using the Solver Profiler.
Solver Resets
The Solver Profiler logs events that cause the solver to reset its parameters because solver resets do incur computational cost.
Zero-Crossing Events
Detect zero-crossing events using Solver Profiler.
Solver Exception Events
This example simulates two identical nonlinear spring-damping systems.
Jacobian Logging and Analysis
The Solver Profiler supports Jacobian logging and analysis for implicit solvers.

Boucles algébriques

Algebraic Loop Concepts
Learn how algebraic loops are created during simulation.
Identify Algebraic Loops in Your Model
If Simulink reports an algebraic loop in your model, the algebraic loop solver may be able to solve the loop.
Remove Algebraic Loops
Learn how to break undesired algebraic loops in a model.
Modeling Considerations with Algebraic Loops
Learn modeling techniques to avoid unnecessary algebraic loops.

Informations connexes

Optimiser les performances
Charger des données de signaux pour la simulation
Diagnostics
Solver Pane
Hardware Implementation Pane

Exemples présentés

Explore Variable-Step Solvers with Stiff Model

The behavior of variable-step solvers in a Foucault pendulum model. Simulink® solvers ode45, ode15s, ode23, and ode23t are used as test cases. Stiff differential equations are used to solve this problem. There is no exact definition of stiffness for equations. Some numerical methods are unstable when used to solve stiff equations and very small step sizes are required to obtain a numerically stable solution to a stiff problem. A stiff problem may have a fast changing component and a slow changing component.

Ouvrir le modèle

Exploring the Solver Jacobian Structure of a Model

The example shows how to use Simulink® to explore the solver Jacobian sparsity pattern, and the connection between the solver Jacobian sparsity pattern and the dependency between components of a physical system. A Simulink model that models the synchronization of three metronomes placed on a free moving base are used.

Ouvrir le modèle

Specify Initial State for Simulation

Simulate a model from an initial state using final states data saved from a prior simulation with and without the operating point.

Ouvrir le modèle

Use Fixed-Step Zero-Crossing Detection for Faster Simulations

Use fixed-step zero-crossing detection to improve simulation performance.

Ouvrir le script

Improve Simulation Performance by Using Local Solvers

Improve simulation performance by using a local solver for a component with much faster dynamics compared to the rest of the system.

Ouvrir le live script