Compressors use a lot of energy to move large quantities of unfinished and finished goods between various process steps and storage. The compressor setup may be efficient when fully used, but may not be used optimally when ramping up and down or under reduced load.
The production is usually controlled by a basic control system (e.g. DCS or SCADA), but as the process needs vary, operation is not always optimized.
SentianController automates the control of the compressor system by learning its system dynamics from historical data, which are then used to predict the optimal settings for the desired goal or combination of goals, even in scenarios that have not been encountered before. At the same time, it reduces variation in output.