Simulink Parameter Estimation 1.2.3

Product Description

• Introduction and Key Features
• Working with Simulink Parameter Estimation
• Working with Measured Data
• Refining Models with Measured Data
• Representing Systems with Adaptive Lookup Tables

Refining Models with Measured Data

The Control and Estimation Tools Manager lets you define and compare estimations and review the results of multiple estimations in one graphical user interface.

Defining Model Variables

Simulink Parameter Estimation automatically detects models with many model parameters and state variables. It lets you select parameters and variables for each estimation task so that you can focus on specific sections of the model. You can use the results of an estimation to reset the values of global parameters and then use these values as starting points in subsequent estimations.



The variables branch of the Control and Estimation Tools Manager lets you specify the parameters used in the estimation. Click on image to see enlarged view.

Estimating Parameters

Within each estimation task, Simulink Parameter Estimation lets you select specific data sets to estimate particular combinations of model parameters and state variables. You can refine your parameter tuning by using parameter values from previous calculations as initial values for subsequent estimations. You can also use these values to set ranges for estimated parameters.



The Control and Estimation Tools Manager helps you select the parameters to be estimated. Click on image to see enlarged view.

Configuration tools let you specify optimization methods and options, such as the number of iterations.Optimization methods for highly nonlinear models include nonlinear least squares and direct search routines. A configurable view of the estimation progress gives you access to step size and other options, letting you inspect the calculation for convergence.

Comparing and Validating Estimations

Simulink Parameter Estimation can generate comparative plots of estimation results to help determine which parameters best describe the behavior of the model. Plots include views of parameter sensitivity, measured vs. simulated model outputs, and residual values. Multiple plots let you compare estimations results generated from different input-output data sets.

Validation involves comparing the model output to an independent set of test data to determine whether the calibrated model accurately represents the system dynamics. Simulink Parameter Estimation lets you compare multiple model outputs to the validation data set to select the best estimation and parameters sets.



You can use plots to validate your estimated results against test data. Click on image to see enlarged view.