With growing fuel prices and global warming, fuel economy
improvement and reduced emissions are becoming order of the day.
Automobile manufacturers around the world are in increasing
pressure to achieve the same, also keeping in account the stiff
timelines required in the product developmental cycle. Condensed
duty cycle that is representative of several days of actual real
life running is developed for quicker fuel economy tuning on a
chassis dynamometer. This paper presents a new methodology to
obtain a synthetic drive duty cycle, which matches engine operating
conditions of the actual real life cycle accurately and thereby
providing a more accurate match in fuel economy.
Drive duty cycle (vehicle velocity profile) used in this study
is extracted from the instrumented vehicle in the real traffic
condition (peak/lean hour) of major cities in India at different
location/load conditions. Each recorded trip is further divided
into micro trips. Several principal components such as average
velocity, percentage of times in acceleration, deceleration, cruise
& idle, shift pattern closeness, (which is a measure of number
of shifts from one of the above-mentioned state to the other),
etc., of the entire trip and individual micro trips are evaluated.
The synthetic drive duty cycle is derived by minimizing the
objective function, which is created from a selected set of the
principal components. An optimization algorithm that reduces the
objective function coded in MATLAB is used to create the synthetic
duty cycle by suitably clubbing the micro trips such that there are
no repetitions. The synthetic cycle and the full trip data are both
analyzed for fuel economy in AVL-CRUISE software. From the
AVL-CRUISE simulation results, fuel economy & engine operating
points of condensed & real-world cycle are compared. Principal
components that yield the closest match not only in fuel economy
but also on engine operating points repeatedly are identified and
presented.
With the identified objective function from a optimal set of
principal components, the engine operating point plots and hence
the fuel economy values matches very well with that of the entire
trip to within ±3% and the total development time is reduced to
less than 1/6th of the time consumed otherwise. This methodology
provides repeatable & reproducible results consistently. Some
of the factors affecting the data obtained from different cycles
are discussed and possible methodologies to develop synthetic
cycles across different vehicle categories/platforms are
suggested.