University & Colleges
Request More Information
Closed Loop System Identification
The PiControl’s Pitops TFI (Transfer Function Identification) is a powerful, multi-input closed-loop system identification tool based on GRG (generalized reduced gradient) nonlinear constrained optimization algorithm making use of 3G (geometric, gradient and gravity) calculation algorithms. The 3G algorithms allow identification of transfer functions even with completely closed-loop data amid noise and unmeasured disturbances. Most other competitor algorithms are susceptible to noise and disturbances, but in contrast, the 3G algorithm inside Pitops TFI can accurately separate the disturbances from the true transfer function dynamics. Read more to find how we can help with closed loop system identification.
The Pitops TFI system identification tool can be even used to teach system identification even at the undergraduate level and community colleges.
The Pitops TFI system identification tool can be used for statistical and math research projects and also in chemical plants, manufacturing processes for system identification (transfer function identification) followed by closed-loop control system design and implementation.
The current ARMAX/ARX etc. approach used in other common competitor products are relatively far more complex and obsolete. Using PiControl’s invention the Pitops TFI system identification algorithm there are the following benefits missing from existing ARMAX methods:
- Bad data splicing not required, noise in data is not an issue, does not require preconditioning.
- Missing data/bad spikes etc. are not an issue- no conditioning required.
- Steps on manipulated variable are not needed- closed loop data can be used, even using just existing historic trend data.
- Detrending and/or normalization of data not required.
- Dead time is automatically estimated, does not have to be set fixed by user.
- Dead time can be as short as milli-seconds or as high as hours and is calculated automatically.
- Multiple inputs can be handled.
- Works well for complex, nonlinear and slow processes with long dead times and long time constants commonly encountered in chemical processes.