With the 2020 EU regulations in sight, hybrid powertrains represent one of the most promising
solutions to reduce fuel consumption and CO2 emission of vehicles. One of the latest ad- vances is
the plug-in hybrid electric vehicle, equipped with an internal combustion engine and a large
battery pack which can be completely discharged during operation. With multiple energy sources
available, the way in which the energy is used is crucial to utilize the full potential. This
report covers an implementation and comparison of two different energy management strate- gies,
performed in two different modeling softwares using a model of the Chevrolet Volt plug-in hybrid.
The powertrain of the Chevrolet Volt mainly consists of a gasoline engine, an electric traction
motor, a generator, a battery pack and a planetary gear set. The way in which the components are
connected allows the powertrain to operate in four modes. The powertrain is modeled in Matlab
Simulink and Gamma Technologies GT-SUITE, using a quasi-static approach. Because the input and
output signals of the energy management are fixed, additional feedback control is needed in the
A heuristic strategy is implemented using a map during electric driving and heuristic rules for
operation of the engine. This strategy is used before at a benchmark competition and achieved a 2nd
place out of 9 competitors. A static optimisation strategy is used in the form of the Equivalent
Consumption Minimization Strategy (ECMS), which is based on optimal control and offers a suboptimal
solution. It is implemented using a matrix of equivalent fuel consumptions, selecting the minimum
value within each mode and between the modes. The adaptation of the equivalence factor is done by
feedback on state of charge.
Simulations are performed on four driving cycles: The NEDC and three real-world driving cy- cles,
one on flat roads and two on mountain roads. The heuristic strategy results in the lowest fuel
consumption on two driving cycles, while the ECMS performs the best on the other driving cycles.
However, the fuel consumption values are still quite far away from the optimal values. When
comparing the two softwares, the fuel consumption during both simulations is similar. The state of
charge profile during electric driving is almost the same, but this changes when the internal
combustion engine is switched on. As a result of using feedback control the torque delivered by the
engine is also slightly different.
Main conclusions are that the ECMS is successfully implemented with the lowest fuel consump- tion
at two out of four driving cycles. The vehicle is modeled in two software environments, using
different approaches of controlling the components but showing comparable simulation results.
Overall a numerical model of the Chevrolet Volt represents an effective tool to asses the per-
formance of different strategies. Recommendations are to further expand the vehicle model
with more detailed components and to change the adaptation of the co-state in the energy