printlogo
http://www.ethz.ch/index_EN
Institute for Dynamic Systems and Control
 
print
  

Energy Management Strategies for Hybrid Electric Vehicles

Project Details

IDSC_logo_Icon


Start Date: May 2005
End Date:   July 2009

 

Contact:

Prof. Lino Guzzella

 

Supervision:

Prof. Lino Guzzella

 

Lead Researcher(s):

Dr. Daniel Ambühl

 

Additional Participants:

New

Press

Video

Links

Publications

A Causal Operation Strategy for Hybrid Electric Vehicles based on Optimal Control Theory,
Ambühl D., Sciarretta A., Onder C.H., Guzzella L., Sterzing S., Mann K., Kraft D., Küsell M., Proceedings of the 4th Symposium on Hybrid Vehicles and Energy Management, Braunschweig, Germany, 2007


On Implementation of Dynamic Programming for Optimal Control Problems with Final State Constraints,
Sundström O., Ambühl D., Guzzella L., Proceedings of Les Rencontres Scientifiques de l’IFP: Advances in Hybrid Powertrains, Rueil-Malmaison, France, 2008


On Implementation of Dynamic Programming for Optimal Control Problems with Final State Constraints,
Sundström O., Ambühl D., Guzzella L., Oil & Gas Science and Technology — Rev. IFP, 2009, Published Online: 25 September 2009, DOI: 10.2516/ogst/2009020

Predictive Reference Signal Generator for Hybrid Electric Vehicles,
Ambühl D., Guzzella L., IEEE Transactions on Vehicular Technology, vol 58, no. 9, 2009

Energy Management Strategies for Hybrid Electric Vehicles,
Ambühl D., Diss. ETH No. 18435, 2009;

Appropriate control of the powerflows in hybrid powertrains is crucial for fully exploiting the fuel saving potential of such configurations. In this project we developed a real-time capable, model-based strategy for fuel optimal control of hybrid powertrains. Depending on the information available onboard, this strategy can be implemented on different levels. The base strategy does not require any information on future driving conditions, whereas two extensions were developed that exploit short and long term information to further decrease the fuel consumption.


This project is finished and led to the dissertation ETH Diss. No. 18435.

Introduction

Hybrid electric vehicles are commonly known as a promising solution to reduce the fuel consumption with existing technology for the near future. Due to the presence of at least two power converters in the powertrain, there is a new degree of freedom compared to conventional vehicles. An appropriate control of this degree of freedom is required to achieve lowest possible fuel consumption.

Objective

The objective of this thesis is to develop novel control algorithms for the before mentioned degree of freedom in order to minimize fuel consumption. These algorithms are evaluated in simulations, but finally, they have to be applicable to real vehicles. Therefore, the algorithms must be causal, which means that they can only exploit information available in the present and the past. Further, they have to be computationally very efficient, since the computational power and the memory capacity in the engine control unit are limited.

Dynamic Programming

This objective is approached by first evaluating the global optimum with dynamic programming as a benchmark for any real-time capable control strategies. Such globally optimal solutions are acausal and can only be evaluated in simulation. In this thesis, a new method is presented that allows to enhance the accuracy and the computational efficiency of dynamic programming for single-state optimal control problems with final state constraints.

Non-Predictive Strategy

Causal control strategies are derived and investigated in a second step. A simplified model for the hybrid powertrain is introduced in addition to the original model. This simplified model allows to derive explicit solutions for the optimal control, resulting in a strategy that is computationally very attractive and allows to gain insight into its structure. An investigation of these causal control strategies with the original and the simplified model has shown, that they achieve very good performance in terms of fuel consumption as long as there are no severe recuperation phases. For driving cycles with elevation profiles, the fuel consumption achieved by these strategies differ significantly from the theoretical optimum.

Predictive Strategies

In order to extend these strategies such that they show good performance even in environments with elevation profiles, some knowledge on the future driving conditions is taken into account. A novel algorithm has been developed in this thesis that evaluates a reference trajectory for the future state-of-charge of the battery such that low fuel consumption can be achieved in conjunction with the previously proposed strategy even in driving conditions with elevation changes. This algorithm exploits data from the navigation system on the trip that is planned such as the topographic profile and the average traveling speeds on each road segment. This predictive algorithm is computationally very efficient and the resulting fuel consumption is improved considerably compared to the non-predictive strategy.

In a last step, the optimal starting and stopping decision for the engine has been investigated. For the case of a full electric hybrid, it is optimal to shut the engine off for some time intervals and to drive electrically. However, if there is no cost in terms of fuel consumption assigned to an engine start, the solutions resulting from optimal control can show frequent starting and stopping of the engine. In order to approach this problem systematically, modeling of the energetic cost assigned to each engine start is required. An investigation of the optimal solution of the energy management problem including starting cost has shown, that this energetic starting cost cannot be neglected for an appropriate control. A model predictive control scheme is introduced for the decision on the engine operation. The performance in terms of fuel consumption of this model predictive control is evaluated as a function of the prediction horizon. These results revealed that a very short prediction horizon is sufficient to achieve close to optimum performance. 

 

Wichtiger Hinweis:
Diese Website wird in älteren Versionen von Netscape ohne graphische Elemente dargestellt. Die Funktionalität der Website ist aber trotzdem gewährleistet. Wenn Sie diese Website regelmässig benutzen, empfehlen wir Ihnen, auf Ihrem Computer einen aktuellen Browser zu installieren. Weitere Informationen finden Sie auf
folgender Seite.

Important Note:
The content in this site is accessible to any browser or Internet device, however, some graphics will display correctly only in the newer versions of Netscape. To get the most out of our site we suggest you upgrade to a newer browser.
More information

© 2014 ETH Zurich | Imprint | Disclaimer | 8 July 2009
top