INCA Discovery: Model Identification for APC
Model predictive controllers, such as INCA MPC, require an accurate model of the process for their optimizations. INCA Discovery is the offline modeling package that complements the online INCA MPC.
All steps relevant to modelling and preparing the model for MPC are done in one place. No other modelling packages are required.
Rich data import and visualization facilities are appreciated even beyond the model identification context.
Process data is imported from a variety of formats, including the INCA Acquire format, csv files and other formats often encountered in MPC context
Add more data later to improve the models without losing results so far
Powerful trending supports the selection of data for the model identification, excluding upsets in the data, etc.
Transformations and user defined calculations assist in selecting data and compensating for non-linearities
Model identification fits the model to the selected data
Constraints on the gains, dead times can be imposed, going beyond the model that just fits the data best
Process knowledge involving several variables can be imposed too, such as gain ratios and mass balance constraints
Flexible post-processing of the models all the way to manually specifying a model gives you all the freedom you need
Analysis of the model conditioning lets you correct problems before they occur online
Understanding the process is key!
Look beyond the controllers. Look at the process, the instruments, the actuators
Keep it simple Organise mainteance. Mainteance is the key to harvest the benefits for many years
Involve operators in the MPC projects
Ready to Maximize Your Asset Performance?