Predictive control for complete vehicle energy management
The purpose of this project is to bridge the gap between Chalmers’ finished projects in Energy Management and an already granted doctoral project which cannot be started until the budget for Phase 3 of SHC is decided.
The project builds on Lars Johannesson Mårdh’s and Nikolce Murgovski’s research on control architecture and algorithms for predictive energy management of conventional and hybrid vehicles, made in the strategic research area (SFO) Chalmers Energy Initiative (CEI). The algorithms have so far only focused on optimizing energy management in a vehicle with information about speed limits and road gradient.
The proposed control architecture and algorithms allow extension with information about the surrounding vehicles and traffic flows, which is the first step toward autonomous driving. However, if the control system should be applied to autonomous vehicles and achieve maximum fuel savings there are many unsolved questions and challenges: which is the best distributed control system for road trains of conventional and hybrid vehicles; how can convex optimization be used to reduce the computational requirements; how can algorithms be extended from mainly motorway driving to city traffic; how should the wear models be included?
A rough plan for the project is to:
1) organize an SHC seminar on CEI control system;
2) expand Chalmers simulation model to manage multiple vehicles;
3) propose a distributed predictive control algorithm for planning kinetic, electrical and thermal buffers.
Ladda ner rapporten: Predictive control for complete vehicle energy management (139 KB)