Control system design starts with an accurate plant model. You can describe the complex dynamics of your plant using a variety of modeling approaches, all supported by MathWorks tools. Because these tools work together in the Simulink® modeling environment, you can use the most appropriate modeling approach for each component in your plant to create the system-level plant model.
When you don’t know the physics of your system, you can employ system identification, a common data-driven modeling technique that uses input-output data to create a plant model. This approach lets you rapidly develop accurate plant models without knowing the detailed structure of the model.
MathWorks physical modeling tools let you construct plant models by using blocks that represent mechanical, electrical, magnetic, hydraulic, pneumatic, and thermal components to map the component topography and physical connections of your system. With this approach, you can efficiently create complex multidomain plant models without having to derive the underlying first-principles equations. If you know the plant equations explicitly, you can implement them by using Simulink blocks or express them in the Simscape™ language. First-principles models—those built with physical modeling tools or by implementing equations explicitly—can be developed without access to physical prototypes or plant hardware.
First-principles models contain parameters that correspond to physical properties of the plant, such as mass, electrical resistance, and flow area, that may not be known. Simulink Design Optimization uses experimental data to improve model accuracy by calibrating model parameters with input-output test data.
Model environmental effects, such as atmosphere, gravity, wind turbulence, and the Earth’s magnetic fields, with Simulink and Aerospace Blockset™. You can create custom environmental models from Simulink blocks, such as lookup tables.
"Without MathWorks tools, we would have had to use C or Fortran to construct a suitable simulation model. We would not have been able to analyze the behavior of the controller and the mechanical system together before prototyping."Read the story