Machine learning, optimization and app development for charging processes of electric vehicles

Develop a user-friendly app that will do optimization support of the charging process of electric vehicles with new, environmental and cost effective methods. The logic will be based on the latest state of art within neural network and with the latest available scalable solutions within the cloud based Google platform.
Master

To develop a user-friendly app that will do optimization support of the charging process of electric vehicles with new, environmental and cost effective methods. The logic will be based on the latest state of art within neural network and with the latest available scalable solutions within the cloud based Google platform.

The app should be made in the Google.cloud platform and programed in Python. The student development team will get 20 000 dollars in credit from Google. The team will test the model and system with two actual cars (Tesla and BMW I3) The result of the master theses should be a downloadable app for iOS and Android.

The students will get close support and tutoring from a former power trader and energy manager from Statkraft with experience in from the power market, optimization and developing of decision tools to the power industry.

Backend, optimization and frontend development will either be available as separate student task or as a group task (2-3 students). The scope and output (work stories) are well defined, but the layout of the backend design can the students influence. The optimization will be developed in Tensorflow with LSTM, recurrent neural network and decision three/forest.

Goal

Develop a user-friendly app that will do optimization support of the charging process of electric vehicles with new, environmental and cost effective methods.

Learning outcome

Depending on the chosen sub-topic, data analysis and processing, machine learning, app development and optimization skills.

Qualifications

Interested in the topic, math, some signal processing and good programming skills recommended but its not a hard requirement.

Supervisors

Collaboration partners

Flexibility AS

References

cloud.google.com/products/
www.tensorflow.org
cloud.google.com/appengine/
cloud.google.com/solutions/mobile/
cloud.google.com/products/machine-learning/