View the Pitops brochure. Click on Schedule a Demo to download and examine a full-blown demo of Pitops-TFI.
MULTIVARIABLE CLOSED-LOOP PROCESS TRANSFER FUNCTION DYNAMICS SYSTEM IDENTIFICATION
Pitops-TFI is a software product for identifying process dynamics (transfer function parameters) using plant data. It provides the following powerful features:
- Transfer function identification using closed-loop data, open-loop data or a mixture of both.
- Simultaneous, mutivariable identification with multi-inputs. Handles both SISO (single-input, single-output) and MIMO (multi-input multi-output) control problems.
- Identifies Control Valve Stiction or Deadband.
- Equipped with a powerful, highly automated, easy to use software interface.
- Runs all in time domain, no complicated discrete (Z) domain knowledge required.
- Equipped with powerful constrained nonlinear optimizer to identify process dynamics
Pitops is the only software product that performs true closed loop system identification with the secondary (slave) PID controllers in Auto mode or even Cascade (Remote) mode. No other competitor tool can do successful transfer function identification using data with PID controllers in Cascade modes (Pitops is the only one). Furthermore, Pitops performs transfer function identification entirely in the time domain whereas all other competitor tools use the more complicated Laplace (S) or Discrete (Z) domain. Pitops can even handle multiple inputs and identify multiple transfer functions simultaneously. Pitops performs multiple input closed loop transfer function system identification in the time domain using a new and novel proprietary breakthrough algorithm that is far superior than the older methods like the ARX/ARMAX/Box and Jenkins methods that are used in competitor tools.
Identify Process Transfer Functions EasilyPitops-TFI provides powerful capabilities for designing and implementing advanced control schemes. It helps to precisely identify process dynamics required for optimizing, tuning of advanced PIDs, feedforwards and model-based control schemes.
Figure 1. Dynamics Estimation for Steam—>Temperature Loop
Anyone Can Use Pitops-TFI (both Engineers and Technicians)Pitops-TFI is so simple that even a new technician or engineer with no prior control knowledge can learn to identify transfer functions using plant data in just about 20 minutes.
Transfer Function Identification Example
Figure 1 shows a transfer function identification between steam flow and distillation column temperature. The red trend with up-and-down steps is the steam flow setpoint. The noisy red trend is the temperature and the blue trend is the transfer function model prediction for temperature.
All Calculations Done in Simple Time Domain
Notice that almost all dynamics identification work can be done simply and easily from this single main Pitops-TFI screen. Since Pitops-TFI works wholly in the time domain, no conversion is needed to go from Z (discrete domain) or S (Laplace domain) to the time domain. Also there is no need to know or understand complicated terms like ARMA, Noise Models, poles or zeros and other complicated academic control jargon. There is also no need to understand cross correlation and auto-correlation theory. Everything is so simple and straightforward for anyone to understand and start using quickly.
What-If Analysis of Transfer Function Process Dynamics
The software also allows you to easily conduct “what-if” simulation studies by specifying guessed values of transfer function parameters and to even compare predicted models with other data sets not used in the dynamic estimation. This feature allows you to quickly and easily check the model accuracy.
Multivariable Closed-Loop Identification using Short Duration Data
Figure 2 shows the simultaneous multivariable closed-loop identification with a remarkably short duration of data. Three input variables (reflux, column feed and reboil duty) shown in the bottom three red trends are changing simultaneously. All three inputs impact the column impurity shown in the top red trend. The time constants for the three inputs are about 45 minutes long. Even though the data window is only about 4 hours long, Pitops-TFI is able to successfully identify all three transfer functions simultaneously. The combined effect of all three transfer functions is the blue trend shown in the top window. Note that this is a powerful illustration showing a true simultaneous, multivariable closed-loop identification.
Control Valve Stiction
Pitops TFI can identify Control Valve Stiction using plant data. The stiction identification location is shown in Figure 3.
Dynamics identification in other competitor products is far more complex, dealing with discrete (Z) transforms. Pitops-TFI makes dynamics identifcation delightfully simple even for a new engineer or technician. Pitops-TFI is the only product required for all possible primary and advanced control dynamics identification in any type of chemical or refining operation.
Click on Schedule a Demo below to download and examine a full-blown demo of Pitops-TFI.
CONTACT US FOR A DEMO