Rule-Based Control of Series HEVs Derived from Deterministic Dynamic Programming
2008 (English)Report (Refereed)
We present a model of a series HEV including resistive losses in the energy storage and varying efficiency of the combustion engine and generator. An optimal control problem minimizing fuel consumption is solved using the model and deterministic dynamic programming. It is evaluated on several standard city bus driving cycles and the solutions are analyzed to extract rules useful for real-time controlling. The resulting real-time strategy is compared to the optimal control using simulations. The strategy has a threshold value determining when to switch off the generator. This threshold value has a driving cycle dependent optimum giving the minimum increase in fuel consumption compared to the optimal control, and the optimal value can thus be chosen based on driving cycle characteristics. We see that the optimal value is related to the mean power demand of the driving cycle. The proposed real-time strategy performs well, and a minimum increase of 5 % can be achieved even on a driving cycle not used for rule extraction.
Place, publisher, year, edition, pages
IdentifiersURN: urn:nbn:se:kth:diva-55910OAI: oai:DiVA.org:kth-55910DiVA: diva2:472109