Discrepancy modeling
Discrepancy modeling
Twin Fabrica implements a proprietary Discrepancy Model, a robust and lightweight neural network designed to guarantee stability, no machine learning expertise required, to refine model precision.
Define 4 simple hyperparameters, and the network can be trained with a single mouse click.
1. Define the required inputs from the measured and estimated quantities.
2. Define the target VTS temperature estimation that you want to improve.
3. Train the model and deploy it within the final hybrid architecture automatically.