Improving motor control algorithms via machine learning at the network edge

The thesis is on how to leverage Machine Learning solutions for improving existing motor control algorithms, on a defined situation which is a Sensorless FOC algorithm based on an Infineon Development Kit. The work that has been achieved consists of: data collection, Machine Learning models implementation, training and fine-tuning, and model deployment on the edge by extracting the model to a C++ library and then integrating it into the C project.