Frequency Domain System Identification Toolbox

Tools for frequency domain model identification and validation

Highlights

  • Design of optimized periodic excitation signals (multisine, binary, three-level)
  • Identification of continuous-time or discrete-time systems with unknown fractional delay
  • Model order selection
  • Model validation, including simulation, calculation of residuals, and test of cost functions
  • Calculation of confidence intervals of amplitude/phase and poles/zeros
  • Model conversion to and from the Control System Toolbox and System Identification Toolbox objects

Description

fdident
Identify the behavior of a robot arm
with an estimate for 4 zeros and 6 poles.

The Frequency Domain System Identification Toolbox (FDIDENT) provides specialized tools for identifying linear dynamic single-input/single-output (SISO) systems from time responses or measurements of the system's frequency response. Frequency domain methods support continuous-time modeling, which can be a powerful and highly accurate complement to the more commonly used discrete-time methods. The methods in the toolbox can be applied to problems such as the modeling of electronic, mechanical, and acoustical systems.

The Frequency Domain System Identification Toolbox is built entirely in MATLAB and all functions are available from the MATLAB command line or through an interactive interface.

For more information, please visit the FDIDENT Developers' Page.

Required Products

Platforms

  • Linux
  • Macintosh
  • UNIX
  • Windows

Support

  • Consulting
  • E-mail
  • Fax
  • On-site assistance
  • Telephone
  • Training

Product Type

  • Data Analysis Tools

Tasks

  • Control Systems
  • Data Analysis and Statistics
  • Vibration Analysis and Control

Industries

  • Automotive
  • Consumer Electronics
  • Industrial Automation and Machinery