Deep Neural Network (DNN) Training Scripts for H2DF Engine Models (alexwin9478/H2dfDnnTrainExpData: Zenodo)

# Deep Neural Network (DNN) Training for H2DF Engine Models

This repository contains MATLAB® code to train deep neural networks (DNNs) / gated recurrent unit (GRU)-based models for the Hydrogen–Diesel Dual-Fuel (H₂DF) project at the University of Alberta (UofA).  
The models are based on experimental data from the 4.5 L Hydrogen–Diesel Engine at the MECE Engine Lab, Edmonton, where pseudo-random binary sequence (PRBS) signals were used to excite the actuators. 

The trained models are later integrated into **nonlinear model predictive control (NMPC)** frameworks for advanced combustion control.  

Please see the ReadMe in the github Repo for more information.