System Identification Toolbox™ estimation, validation, and analytical functions accept input/output estimation data in multiple forms.
Time-domain data — Timetables, numeric matrices, time-domain
Frequency-domain and frequency response data — Frequency-domain
You can also generate custom signal data to provide the stimulation signal for experiments or to study estimated model behavior by simulating the model response to these signals.
Data types such as timetables and
iddata objects also include
properties that include information about the data such as sample rate, units, and, for
iddata objects, intersample behavior, channel names, and experiment
identifiers. Numeric matrices contain only data values, and provide no information on sample
rate or any other data properties.
You can combine related data sets as long as they have the same sample rate and channel selections. In particular, you can create multiexperiment data sets that must share sample rate and channel selections but can have different durations and start times.
Use Different Data Types
|Tables for time series data, with timestamped rows and variables of different types|
|Input-output data and its properties for system identification in the time or frequency domain|
|Frequency response data or model|
|Convert time-domain |
Generate Custom Signal Data
- Representing Data in MATLAB Workspace
Represent time-domain, time-series, and frequency-domain data.
Work with Data Types
- Data Domains and Data Types in System Identification Toolbox
System Identification Toolbox accepts timetables, numeric matrices, and data objects for model estimation in the time and frequency domains.
- Use Timetable Data for Time-Domain System Identification
Create and use timetables for model estimation.
- Use Matrix-Based Data for Time-Domain System Identification
Use data contained in numeric matrices for time-domain model estimation.
- Convert SISO Matrix Data to Timetable
Convert matrix-based SISO estimation data to timetables for model identification.
- Convert MIMO Matrix Data to Timetable for Continuous-Time Model Estimation
Estimate a continuous-time MIMO model by first converting matrix-based data to a timetable.
- Representing Time- and Frequency-Domain Data Using iddata Objects
iddataconstructor to represent time-domain and frequency-domain data and working with
- Managing iddata Objects
iddataobject stores time-domain data or frequency-domain data and has several properties that specify the time or frequency values.
- Representing Frequency-Response Data Using idfrd Objects
idfrdconstructor to represent frequency-response data and working with
Generate Input and Output Data
- Generate Data Using Simulation
Creating input data with specific characteristics and simulating the output data from a model.
Work with Data in the App
- Import Time-Domain Data into the App
Import time-domain data into the System Identification app.
- Import Frequency-Domain Data into the App
Import frequency-domain input-output data and frequency-response data into the System Identification app.
- Import Data Objects into the App
- Specifying the Data Sample Time
Specify time between successive data samples.
- Managing Data in the App
You can get information about each data set in the System Identification app by right-clicking the corresponding data icon.
Use Complex-Valued Data
- Manipulating Complex-Valued Data
Supported operations and limitations for handling complex data and commands for manipulating complex signals.