Energy Management in a Hybrid Vehicle Using Predicted Road Slope Information

Abstract

Energy management in a hybrid vehicle is about optimizing a problem with several uncertainties. By decreasing the number of uncertainties new methods can help to provide better strategies. In a non-predictive energy management, there is no information available about the future driving conditions whereas in a predictive method future information together with past and present data is used within the strategy. If all future information is available, the optimal solution for the drive cycle can be found, but in reality, due to uncertainties in the driving conditions, some estimations must be made. Therefore it is interesting to see how limited information can affect the fuel consumption. In this thesis, a method is developed based on having limited information about the upcoming driving condition consisting of an altitude profile provided by GPS, but there is no information about future speed profile. The results show up to roughly 2.5% less fuel consumption in comparison with a non-predictive energy management strategy, depending on the drive cycle. While fuel consumption is a decisive factor to conclude in an energy management strategy, battery wearing and drivability are inevitably needed to be considered as well. The effectiveness of this new method depends on the level of variations in altitude profile and capacity of the battery. In more hilly drive cycles there is more opportunity to employ the information provided by altitude prediction of the road. Furthermore, it is showed that increasing the battery size can help the vehicle to have better performance in very hilly roads. As a result in one of the hilly drive cycles 100% increase in the battery capacity yield 2.7% less fuel consumption whereas in case of halving the battery size the increase in the fuel consumption is 6.5%.