This project aims to apply Data analytics / Machine learning to optimizing the production of liquefied Natural gas from an existing plant.
Real Industrial data with about 70,000 records of activities in the plant was analyzed to build a model that predicts the LNG flow.
Then, an algorithm was created to interact with the model, using a form of binary search technique to find the best input variables that can optimize a particular state in the plant.
With the model, we were able to achieve an average of 10% increase in the optimization of the LNG production.
A Graphic user interface was designed to give a visual interaction with the work that has been done.
This work was done in an anaconda environment.
To use the app: Run
python gui.py
