MPC based Transfer learning for HCCl Optimization

The growing demand for mobility worldwide and the mass use of internal combustion engine drive systems poses major challenges for today’s society. Currently, the CO2 content in the atmosphere is continually increasing and the concentrations of pollutants in many areas exceed the maximum permissible limits in many urban centers. Additionally, these engine emissions contribute significantly to global warming. The proposed research looks to reduce these engine out emissions with the implementation of optimal control, in this case model predictive control (MPC). The research will focus on designing and implementing an MPC controller at RWTH university on a single cylinder research engine. The controller will be tested at various operating conditions. The second phase of the proposed project is the transfer learning of the engine controller that will allow the controller developed in Germany to be used on a different engine in Canada.

Faculty Supervisor:

Charles Robert Koch

Student:

Partner:

Rheinisch-Westfälische Technische Hochschule Aachen

Discipline:

Engineering

Sector:

Education

University:

University of Alberta

Program: