Hybrid Powertrain Control A Predictive Real-time Energy Management System for a Parallel Hybrid Electric Vehicle Master of Science Thesis


The purpose of this diploma work has been to see whether it is possible to develop a rule-based controller that mimics the behavior of an optimal control strategy for a hybrid city bus. This control strategy improves fuel efficiency by use of preview information about the road ahead. The rule based controller has been designed for easy implementation into the ISAM engine management system. Dynamic programming is used to find an optimal solution which in turn is used as a blueprint for a rule-based controller. The transition from optimal control to rule-based control is carried out using fuzzy logic. Preview information will in reality be given from a topographic map combined with a GPS. This information together with a speed curve will give information about the future power demand. To simulate a system like this we use parts of the drive cycle in front of the vehicle, which gives both the road incline and the desired speed. Our simulations show that the fuel reduction on a city bus route is about 3.5% when using optimal control, compared with the ISAM control system which is used as a reference system. When the rule-based control is used, the fuel reduction is about 2%. These results have been obtained by controlling the torque split between the internal combustion engine and the electrical machine, without optimizing gear selection. We also carried out simulations including optimization of gear shifting. This resulted in a fuel reduction of about 12%. However, these results are based on somewhat unrealistic presumptions, i.e. gear shifting occurs instantaneously. They are therefore not considered in the rule-based controller. PREFACE This project is a diploma thesis for Master of Science degree from the Electric Engineering program at Chalmers University of Technology in Gothenburg. The project has been carried out at Volvo Technology Corporation. The problem we have worked with is complex and standard control strategies are insufficient. Model-based controllers could be an option, but since the system is of such complexity and the time of a thesis work is limited we have tried a different approach. The method we have chosen is called Fuzzy Logic, and has been adopted extensively in Asia and other parts of the world. The technique is well suited to mimic the behavior of a tutor, or optimal solution. It has been interesting to both work full time in an industrial project and to learn new methods. All the …

40 Figures and Tables

Cite this paper

@inproceedings{Persson2008HybridPC, title={Hybrid Powertrain Control A Predictive Real-time Energy Management System for a Parallel Hybrid Electric Vehicle Master of Science Thesis}, author={Joakim Persson and Thomas Lundberg}, year={2008} }