Information Technology Reference
In-Depth Information
acceleration phases. Compared to automotive applications, the great advan-
tage during the development of operating strategies for hybrid railway vehi-
cles is the fact that the duty cycles are known beforehand. Thereby, optimal
operating modes can be computed almost completely oine. Hillmansen and
Roberts carried out a kinematic analysis of energy storage systems which
suggests a potential of up to 35% energy savings for commuter vehicles [1].
The result of the research work on hybrid concepts for diesel multiple units
in [4] suggests a potential for reduction of fuel consumption of up to 25%
on a fixed route. In [3], a description of modern onboard storage technolo-
gies is given. These are flywheel systems, hydrostatic accumulators, double
layer capacitors, and batteries. All predictions of energy saving potentials in
the above-mentioned studies arise out of simulation studies, a combination
of modeling and optimization is not yet state of the art.
In this paper, a simulation structure for a parallel hybrid railway vehicle
is shown in Sec. 2. An exemplary mathematical modeling is given for the
vehicle. In Sec. 3, an optimality criterion is formulated by a parameteriz-
able performance index quantifying fuel consumption and emissions. Sec. 4
presents the simulation and optimization results. Finally, Sec. 5 concludes
this paper and gives an outlook on future research.
2
Modeling of the Basic Parallel Hybrid Structure
The power train of a basic parallel hybrid railway vehicle mainly consists of an
internal combustion engine, an electric motor, and an energy storage device.
Fig. 1 shows the schematic block diagram of the parallel hybrid system in
which the blocks visualize the models of the hybrid system components. Here,
all main effects contributing to the longitudinal dynamics were modeled. To
build up a complete system model, consistent interface variables representing
the flow of power and energy were defined for each component. Hence, other
structures, for instance, a serial hybrid structure, can be built up easily by
rearranging the components of the drive chain. In principle, two alternative
calculation approaches are possible for simulation studies of hybrid power
trains: either the forward or the backward calculation. The dashed arrows
in Fig. 1 stand for the forward calculation, where the main input is the
engine torque commanded by the driver. The remaining system variables
are determined by numerical integration as indicated by the dashed arrows.
Note that these arrows represent the order of the calculation steps and not
the direction of power flows.
The backward approach corresponding to the solid arrows, on the con-
trary, represents the solution to an inverse problem. Here, all system variables
are derived from the duty cycle subject to the chosen operating strategy.
In the optimization, only the backward approach is employed. For a given
duty cycle, the necessary drive torque is calculated in the vehicle subsystem.
The resulting power is split after the planetary gear box between the electric
motor/generator and the internal combustion engine. The output is given
